MIME-Version: 1.0 Content-Type: multipart/related; boundary="----=_NextPart_01C52880.DC295B70" This document is a Single File Web Page, also known as a Web Archive file. If you are seeing this message, your browser or editor doesn't support Web Archive files. Please download a browser that supports Web Archive, such as Microsoft Internet Explorer. ------=_NextPart_01C52880.DC295B70 Content-Location: file:///C:/C9344E54/Paper14.htm Content-Transfer-Encoding: quoted-printable Content-Type: text/html; charset="us-ascii" Documentation Costs Avoided

Documentation Costs Avoided

using Python and other Open Standar= ds

 

Andrew Jonathan Fine

 

Operating Systems Software Organization

Engines, Systems, and Services

Honeywell International

 

 

The purpose of this paper is to show how I integrated Python, COM, DocBook, Op= enJade, and Word together to create a documentation tool to support BEACON, the cor= e of my organization’s visual programming  environment.

 

When I started out, I did not have any experience at SGML, XML, or other markup languages.= It was only when I was perilously close to reinventing the concept of markup languages, by adapting direct PythonCOM to Word interfaces from Mark Hammon= d's book, Python Programming on Win32, that I decided there had be a better way to do this from a design and maintenance standpoint.<= /p>

 

A Web search provided me with a crash education on markup languages such as XML, HTML, a= nd SGML  My search also provided insight on how DocBook SGML was an overwhelmingly popular open standard, and on how an open source package named OpenJade could translate a DocBook SGML file into Word Rich Text file.

The arrangement wasn't perfect but I realized: <= /o:p>

·        I could save a year in development and maintenance.

·        I could respond rapidly to new assignments.

 

Throughout this paper and its appendices I have inclu= ded data, listings, and references to downloadable public DocBook DSSSL archive= s.

 

Together they should work sufficiently well on a Win32 workstation for a reader to understand the basic principles under discussio= n.

 

 

 

The Core Data = Flow

 

My mission was to translate arbitrary data mined from various sources scattered throughout my local organization; into sensible-looking Microsoft Word 97 reports.

 

Early on I decided that this was best handled through= a core pipeline of applications that cooperated with each other using common = data conventions, such as in Figure 1. 

 

It seemed sensible to have a single Python generator application for controlling a set of front end translators, a content inser= ter, and a post-processing formatter

 

The generator application would decide report layout,= and would be designed as a class to help cope with requests for new types of reports.   The front end translators arrange data (pictures, tables, paragraphs) into a dictionary. =

 

The inserter would create a Word document and insert items into the document selected by the generator.   A post-processing formatter = takes this and modifies it according to the latest corporate Word format style template.

 

The flow was designed to cope with changes in requirements by the different teams within our department, and by corporate-level standards.

 

Figure 1: 

Original Core Data Flow<= /span>

 

The first front-end translator I needed to create was= to take pictures, tables, and data from a recursive property list constructed = by an aerospace industry software visual programming tool called BEACON.  

 

The actual design of this translator is beyond the sc= ope of this paper, however, from this translator I was able to obtain a huge am= ount of sample data fit for testing an inserter.

 

The design of the Word content inserter was based on a straightforward use of principles demonstrated by Mark Hammond's book, Python Programming in Win= 32.   There is a chapter containin= g a thorough treatment of how to have Python use the Word 97 COM object model to create = and manipulate a Word Document.

&nbs= p;

From this experience, I will state that one of the nicest things about Python is its win32com library.  With it, the majority of Office Automation chores reasonably expected of a Visual Basic programmer can be done without needing to buy Visual Basic or related programming tools, debuggers, or libraries.

&nbs= p;

I would highly recommend Python over Visual Basic to anyone contemplating the need = for maintainable and scalable Win32 GUI and batch applications containing the maximum fraction of code that can be redeployed to other uses.  This is especially true for those manipulating COM interfaces.

Problems with the Inserter

&nbs= p;

I initial= ly took the direct approach to inserting and specifying the content in specifi= ed context using specified style.  This involved writing a tree of class methods where each method invoked and chec= ked results of a COM interface.

&nbs= p;

I started confronting difficulties with this approach after a month of going down this design path.

&nbs= p;

It turned= out that the speed of the COM interface was unacceptably slow when inserting th= e massive amounts of table cells that I had culled from the BEACON source code. =

&nbs= p;

Worse sti= ll were the reuse issues that I was beginning to confront with the classes I w= as writing for dealing with the different stylistic issues at the different le= vels of sections, headings, paragraphs, and phrases.

&nbs= p;

To address the reuse issue, I was actually beginning to consider writing ASCII text fi= les for specifying margins, font, heading level, and insertion points for stand= ard forms.  I was starting to get overwhelmed by the design issues after spending a week on this. 

&nbs= p;

I was now perilously close to designing my own standard for text-specified typesetting.   If I went forward I would become responsible for maintaining this standard as well as= any code complying with it for as long as I stayed at my department.  Did I really want to do this?  No.

&nbs= p;

My depart= ment wanted to use Python as more than simply a clearer and more composable substitute to subsume our existing Perl, MSDOS, and in-house direct I/O batching language for control system test stands. Inventing standards and solutions in-house for common problems had long ceased to be cost-effective= for us, we very much needed to take advantage of what other people had done.

&nbs= p;

Python ha= s such a tremendous user base that extra Python libraries existed on the Web to provide access to already existing open standards and the open source syste= ms that implement them.   

&nbs= p;

Ideally, = I wanted to find a Python API that would quickly generate nicely typeset copy in Wor= d 97 for a limited set of documents, but in 2001 there was nothing available, and after a month's worth of work this problem had become non-trivial.

&nbs= p;

The next question to ask was this:  was= there an existing open standard for typesetting that a Python API could be quickly written to wrap around?


Shopping for a Standard

 

I spent a week surveying on the Web what the rest of the world had done for open sour= ce automated typesetting.  Arguably I shoul= d have done this before writing a single line of code, but this was my first emplo= yment where I was allowed extensive Web access from my desk's computer.  

&nbs= p;

I did fin= d two very popular standards for transforming ASCII text into typeset copy:  TeX and DocBook.  Both of these standards had open source implementations in C.

&nbs= p;

I wanted = to choose the standard that would best fit a simple Python API.  Ideally, a document should be gene= rated from a tree of method calls where each call mapped to a different level of text.

&nbs= p;

DocBook h= ad more clearly defined and documented production rules for typesetting elemen= ts.  DocBook, The Definitive G= uide is a real treasure when it comes to explaining these rules in detail.<= /o:p>

&nbs= p;

DocBook specification text has two flavors:  SGML and XML.  They bot= h have nested structures of tags with either options or enclosed text.  

&nbs= p;

The XML s= tandard was still under development during 2001, so that left SGML as an available reliable standard.  An open so= urce rendering engine existed to translate SGML directly into Microsoft Rich Text Format.  From there, I could h= ave Python control Word to load the content as RTF, perform any necessary postprocessing cleanup, then emit the final Word document file.<= /span>

&nbs= p;

These findings changed the overall system architecture to that shown in Figure 2.=

Figure 2:

Revised System Architecture using Python and DocBook

I also fo= und a webpage t= hat was extremely helpful for aligning the contents of DocBook download components = into a workable hierarchy.  A refor= matted copy of this webpage in included as Appendix A.

&nbs= p;

I downloa= ded the components shown in Table 1 based on the advice of that webpage:

 

 

Download=

Description

Jadew1_= 2_1.zip

OpenJad= e, an open source DSSSL rendering engine

docbk41= .zip

DSSSL f= iles representing DocBook SGML 4.1 standard

docbkx4= 12.zip

DSSSL f= iles representing DocBook XML  4.= 1.2 standard

Docbook= -dsssl-1.73.zip

DSSSL f= iles containing core DocBook definition

ISOEnts= .zip

Additio= nal DSSSL files for special fonts and symbols

&nbs= p;

Table 1: 

Components for using DocBook on Win= 32 platforms

 

&nbs= p;

The webpa= ge instructions were tedious and did not precisely fit my situation, so I wrot= e a Python script to do the work for me.  This script is included as Appendix B.

 

The files listed in Table 2 provided some very useful= documentation in support of the above downloads:

 

 

Download=

Description

tdg-en-= html-2.0.2.zip

DocBook: the Definitive Guide, English edition.<= span style=3D'mso-spacerun:yes'> 

Docbook= -dsssl-doc-1.73.zip

Documen= tation on DSSSL files for code DocBook definition.

 

Table 2: 

Recommended DocBook documentation

 

Finally, I had needed to make some special modificati= ons to OpenJade and DocBook to provide extra features that various teams wanted= , these are included as Appendix C, D and E.

An Example Python API for DocBook<= /p>

 

DocBook: The Definitive G= uide provided me with enough background knowledge to enable me to design a Python API to DocBook.

&nbs= p;

As an example, suppose we want to generate the&n= bsp; figure as part of a Word document:

&nbs= p;

Name=

Type

statex

Integer

statey

Long

&nbs= p;

Figure 3:

A DocBook informal table rendered by OpenJade into Word

&nbs= p;

&nbs= p;

Using loc= al DocBook definitions this would look like Figure 4.

&nbs= p;

<!DOCTYPE informaltable SYSTEM "C:\Local.dtd">

<informaltable frame=3D'all'&g= t;

<tgroup cols=3D'2' colsep=3D'1' rowsep=3D'1' align=3D'center'>

<colspec colname=3D'Name' colw= idth=3D'75' align=3D'left'></colspec>

<colspec colname=3D'Type' colwidth=3D'64' align=3D'center'></colspec>

<thead>

<row>

<entry><emphasis role=3D'bold'>Name</emphasis></entry>

<entry><emphasis role=3D'bold'>Type</emphasis></entry>

</row>

</thead><= /p>

<tbody>

<row>

<entry><phrase role=3D'x= e' condition=3D'italic'>statex</phrase></entry>

<entry>Integer</entry>= ;

</row>

<row>

<entry><phrase role=3D'x= e' condition=3D'italic'>statey</phrase></entry>

<entry>Long</entry>

</row>

</tbody><= /p>

</tgroup>=

</informaltable>=

 

Figure 4:

DocBook SGML sample used as input to OpenJade for generating Figure 3

 

To write the application in Python, we can use a subset of the DocBook SGML productions, implemented as a Python class in Appendix F.=

 

This class contains those DocBook definitions needed to implement the above example.  The Python code star= ting on the next page shows how this class is used as a library to generate the above DocBOOK SGML.

 

3D"Text

Figure 5:=

Python code for generating DocBook SGML sample in Figure 4

<= ins cite=3D"mailto:Honeywell" datetime=3D"2005-03-14T10:28"> 

Some explanation is ne= eded.  Class DocBook from DocBook.py in Appendix F is the top-level interface callable class.   All that is needed to write a DocBook application is a local class variant that inherits from class DocBo= ok, such as class Example.

 

The inheriting class (Example) overrides methods= and data from the class DocBook to specify custom behavior:

 

·        The class instance attribute, self.data, is overriden at runtime to specify our= own local class variant, InformalTable, of DocBook's informal table class, DocBook.Rules.InformalTable.

 

·        The class data attribute, SECTION, is overridden to specify the title of our lo= cal InformalTable.

 

In turn, our own local variant classes inherit f= rom other DocBook interface callable classes, and override class or instance da= ta as needed to provide an unbroken chain of calls to generate the SGML.<= /o:p>

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

The OpenJade Interface

 

The DocBook package is an extensive set of directories of files written in DSSSL, the Document Style Semantics and Specification Language.  DSSSL= is a special variant of Scheme, an open-source implementation of the Lisp knowledge-based programming language.

 

 DocBook DSSSL specifies the DocBook style of text translation from SGML or XML into low-level DSSSL semantics.

 

OpenJade = is an open source DSSSL execution engine available from SourceForge (www.sourceforge.net). It reads the = DocBook DSSSL stylesheets and  user's = local DSSSL stylesheets if any.  The= DSSSL is executed upon the user's SGML source text to write a final document to l= oad into the user's word processor.

&nbs= p;

In our ca= se, we want to automatically generate files readable by Microsoft Word, so I ha= ve OpenJade set to emit Microsoft Word Rich Text files.

&nbs= p;

OpenJade = is operated as a command-line application, as such, it is simple to design Pyt= hon code to run it from a Popen4 Python standard library call.

&nbs= p;

Assuming a proper DocBook DSSSL hierarchy has been laid out for OpenJade (Appendix A-E) with environment variables properly set, one can execute OpenJade as follow= s:

&nbs= p;

jade -E4000 -t rtf -d Local.dsl c:/usr/sgml/jade/xml.dcl c:\docbook.sgml=

 

Where:

&nbs= p;

·        -t selects the final document forma= t

·        -d selects our local DSSSL styleshe= et

·        the next argument specifies the top-level DSSSL declaration

·        the final argument specifies a file containing DocBook SGML (such as from Figure 3)

 

Because of differing versions between DocBook DSSSL, other archives, and OpenJade, war= ning or minor error messages usually appear having no significant effect on the appearance of the final document. &nb= sp;

 

I do test= the SGML text output by the Python DocBook classes against an unfiltered invoka= tion of OpenJade, to confirm the pattern's overall integrity.  Once the classes are in a producti= on environment, I then set the -E switch to have OpenJade ignore diagnostics resulting from large amounts of SGML input.

Post-Processing using Word Automation with PythonCOM

 

The Micro= soft Rich Text Format files created by OpenJade are quite attractive in overall appearance.  However, this appearance did not conform with many of the corporate level standards for formatted documentation in Microsoft Word Document files.  A local DSSSL stylesheet was writt= en to override several of the default DocBook DSSSL settings.

&nbs= p;

This did = not address how most of the corporate level standards involved Microsoft Word s= tyle identifier names that the generated documents had to point to. 

&nbs= p;

To address this, reformatter is needed as the final s= tage of the document pipeline. It traverses the table, figure, heading, and sect= ion level style identifiers at the various levels of the generated RTF document= 's COM object model.

 

It renames style identifier names to conform with the actual ones used by a copy of Microsoft Word Document Template (DOT) file handed out as a standard by our local reprographics department.<= /span>

 

Finally, it saves the modified document as a Microsoft Word document.

 

None of these tasks were particularly difficult.  Once the COM interface to an Win32= application supporting same is well understood, that application devolves to just anoth= er library in the hands of a Python designer.

Mission Accomplished

 

The assumptions to derive return on investment are conservative.  They may be hig= her due to other factors explained later.

 

I spent the bulk of 2001 developing a system using the ideas in this paper to translate content from a BEACON visual programming language file directly into a Word document in a completely hands off manner.  In 2002 I also perfor= med significant revisions.  My tot= al effort in development, maintenance, and support was about half time over a = two year period.

 

Between the years 2002 and 2003, my department had 5 = ongoing projects at various stages of development ranging in complexity from 30 vis= ual programming files to as many as 150, perhaps 75 on average. 

 

During each of these years for each of those projects, there were at least 2 major mandated releases where the important contents = of every file had to be peer reviewed:  examined in detail by no less than 3 engineers simultaneously (moderator, author, and inspector). 

 

Each of these releases required every visual programm= ing file be rendered into a viewable hard copy format containing all its diagra= m; and a cross-referenced table of all identifiers in every diagram with stora= ge classes, ranges, initial values, documentation, and other fields.

 

The visual programming language GUI application, BEAC= ON, had no comprehensive hardcopy generator.&n= bsp; Instead, it would take an entry level engineer working under moderate supervision to inspect the file with BEACON running on UNIX over a UNIX to Win32 X terminal emulator, manually transferring text from the X terminal i= nto an open Word document.

 

The least complex of these files (about 1/5) would ta= ke half a day.  The bulk of the f= iles (3/5) would take an entire day on average.=   The most complex of these files (about 1/5) would at least 2 days.

 

This stood to waste significant engineering labor that was better spent in improving the quality of my department's software produ= cts.

 

   &nb= sp; Each project release:   1/5 * 75 *  4 hours  =3D     60 hours

   &nb= sp;            =              3/5 * 75 *  8 hours  =3D    360 hours

   &nb= sp;            =              1/5 * 75 * 16 hours  =3D    240 hours

        =             &nb= sp;                      =         =       -----

        =             &nb= sp;                =                 =       660 hours

 

   &n= bsp; Two major releases per year:          * 2     =3D  1,320 hours

     Five projects needing r= eleases:       * 5<= span style=3D'mso-spacerun:yes'>     =3D  6,600 hours

   &n= bsp; Two year period (2002-2003)           * 2     =3D 13,200 hours=

        =             &nb= sp;            =             &nb= sp;       ------

    <= /span> Total effort avoided:        =             &nb= sp;      13,200 hours

 

Once automated without major problems, the generation= of hardcopy for each project would tie up a single person to babysit the proce= ss for up to an entire day, worst case.

 

   &n= bsp; Automated releases over 2 year period:        =      160 hours

   &n= bsp; My effort (12 * 140 hours per labor month):      1 680 hour= s

   &n= bsp; Total investment:        =                         1 840= hours

 

Net effort avo= ided, 2002-3:                    =       11,360 hours

Net avoided by customers 2002-3 at $100/hour:     1,136,000 dollar= s

 

Net labor years avoided 2002-3 at 1680 hours/year:  <= /span>   6.76 years<= /p>

Headcount avoi= ded per year:        =             &nb= sp;       3.38 people  

 

ROI (Total effort avoided / total invested) 2002-3:    7.17

        =       

Clearly, the return on investment (ROI) for automating the generation of documentation just for formal releases to customers clear= ly helped my department avoid a substantial amount of manual labor cost. =

 

Sometimes a new capability provides new conveniences = that engineers can exploit to increase the pace or quality of a project, saving = more money in ways too hard to measure directly.

 

For example, before this capability was in place, a p= eer review meeting involving BEACON usually had to take place in front of an author, moderator, or inspector's terminal.  Now, engineers are more often able= to convene in an actual meeting room, since now it becomes convenient to gener= ate hardcopy for BEACON files under inspection just prior to the meeting.<= /o:p>

 

One beneficial side effect of designing an automated application for this purpose is being able to write a utility to automatica= lly install and integrate subcomponents. 

 

It sometimes took hours for a engineer to manually customize his or her UNIX account to use BEACON from an NT workstation, set= up the X terminal and remote shell interfaces, and cross-mount the local NT dr= ive so that the UNIX account can see it. =   An automated setup process for all this takes now less than 5 minute= s to execute.

 

That was a spinoff of Python code to launch BEACON fr= om NT under UNIX for BEACON to generate pictures to include in DocBook SGML fi= les.

 

The Road Not T= ravelled

 

Finally, we could consider the cost of using other technologies. 

 

The two most popular programming languages that are u= sed for Office Automation are C++ and Visual Basic. 

 

If C++ were used rather than Python, I would need to build or find open libraries for COM control and SGML generation. To integr= ate these libraries would have required, for me, extensive debug pointers and buffers and would have required at least double the effort.

 

I have do= ne created GUI’s and Office Automation in Visual Basic for a previous employer.<= span style=3D'mso-spacerun:yes'>   Visual BASIC does not offer = the developer true object-oriented constructions for inheritance, polymorphism,= and overrideable methods and data.  These limitations greatly impact the development time. 

&nbs= p;

Because o= f my knowledge level of C++, my efficiency across the development cycle would ha= ve been half that of developing in Python.&nb= sp; Because of the technical limitations of Visual Basic, it would have = been reduced to a third or worse.

&nbs= p;

The other possibility is choosing to reinvent a typesetting standard in terms of Pyth= on and COM rather than taking one off the shelf to integrate with Python. 

 

The issue= s I had confront after a single month of effort proved to be daunting.  Were I forced to rimplement even th= at fraction of the features actually needed from DocBook, I would have needed 5 more months to finish development (6 total), with same in maintenance.=

 

Python and DocBook together proved to be a formidable combination for eliminating a real-world business process bottleneck.&nb= sp;

&nbs= p;

The decis= ion of my department to adopt Python and to allow me to use it along with anoth= er open standard, DocBook, has been vindicated by a substantial return on inve= stment over a medium term period of time, even if only in terms of documentation c= osts avoided.

 

Appendix A

Configuration of DocBook downloads for Win32<= /span>

&nbs= p;

http://lists.oasis-open.org/archives/docbook-apps/200011/msg00183.html

&nbs= p;

docbook-apps message

Subject: Re: DOCBOOK-APPS: Transforming DocBook/XML with jad= e on Win32

·         From: Rune Enggaard Jensen <r.e.jensen@bigfoot.com>

·         To: docbook-apps@lists.oasis-open.o= rg

·         Date: Wed, 29 Nov 2000 16:33:37 +01= 00 

 
 
OK, thi=
s is it! I have everything working with no errors now. In this posting I wi=
ll describe (very) briefly what I did. Later I will make a more elaborate d=
escription and submit it to the FAQ. I use Windoze NT 4.0, DocBook/SGML 4.1=
, DocBook/XML 4.1.2, DSSSL 1.59 and jade 1.2.1/openjade 1.3.  Furthermore I use a set of SGML ISO entitie=
s found at Oasis (http://www.oasis-open.org/cover/ISOEnts.zip).
 
In the following I have assumed th=
at you have already installed the packages mentioned above.
 
Here is my placement of the files:=

 
NOTE: I=
 have moved the XML versions of the ISO entities from the directory ....\do=
cbook\xml\ent to the directory ....\ISOentities\XML
= 
 
 

 
C:\usr\packages\sgml
  +--DocBook=

  |  +--SGML
  |  |  =
+--4.1
  |  +--XML
  |     +--4.1.2
  +--DSSSL
  |  +--ALOC
  |  |  =
+--IDEAS
  |  +--DocBook
  |     +--bin
=

  |     +--common
  |     +--contrib
  |     |  +--html
  |     |  +--imagemap
  |     |  +--print
  |     |  +--renumberinpart
  |     |  +--subdoc
  |     +--doc
=

  |     |  +--html
  |     |  +--lib
  |     |  +--print
  |     |  +--testdata
  |     +--docsrc
  |     |  +--htmlpr
  |     |  +--libref
  |     |  +--printpr
  |     +--dtds
  |     |  +--dbdsssl
  |     |  +--decls
  |     |  +--html
  |     |  +--imagelib
  |     |  +--olink
  |     +--frames
  |     +--html
  |     +--images
  |     |  +--callouts
  |     +--lib
=

  |     +--olink
  |     +--print
  |     +--test
  |        +--cases
  |        +--imagelib<=
o:p>
  |        +--xml<=
/o:p>
  +--ISOentities
  |  +--SGML
  |  +--XML
  +--jade
  +--OpenJade
     +--=
bin
     +--=
dsssl
     +--=
jadedoc
     |  +--images
     +--=
pubtext

 
I have a centralized catalog in the top SGML dire=
ctory containing the
lines
 
  CATALOG c:/usr/packages/SGML/jade/catalog
  CATALOG c:/usr/packages/SGML/DSSSL/docbook/catalog<=
/o:p>
  CATALOG c:/usr/packages/SGML/DocBook/SGML/4.1/docbook.ca=
t
  CATALOG c:/usr/packages/SGML/DocBook/XML/4.1.2/docbook.c=
at
 
If you use openjade, you must change the first line to
 
&=
nbsp; CATALOG c:/usr/packages/SGML/openjade/dsssl/catalog=

 
I have set the following environme=
nt variables:
 
&=
nbsp; SGML_CATALOG_FILES=
=3Dc:/usr/packages/SGML/catalog
  SP_CHARSET_FIXED=3Dyes
 
I have not set the SP_ENCODING var=
iable, since this is only necessary
when using the unicode-encoding (correct?).
<=
pre
style=3D'line-height:8.0pt;mso-line-height-rule:exactly'> 
I have edited docbook.cat for the =
XML version: I added the lines
 
&=
nbsp; OVERRIDE YES
  SYSTEM "urn:x-oa=
sis:docbook-xml-v4.1.2" "docbookx.dtd"
 
just before the line
 
&=
nbsp; PUBLIC "-//OASIS//DTD DocBook XML V4.1.2//EN" "=
docbookx.dtd"
 
so I can use the method suggested =
by Norm in his column "If You Can Name
It, You Can Claim It!". See
 
http://www.arbortext.com/=
Think_Tank/Norm_s_Column/Issue_three/issue_three.html
 
I have edited the public identifiers for the ISO entities from the f=
orm
 
&=
nbsp; PUBLIC "ISO 8879:1986//ENTITIES Diacritical Marks//EN&quo=
t;
      =
;   "ent/iso-dia.ent"
 
to the form
 
&=
nbsp; PUBLIC "ISO 8879:1986//ENTITIES Diacritical Marks//EN//XM=
L"
      =
;   "../../../ISOentities/XML/iso-dia.ent"<=
/o:p>
 
Note that two things were changed =
in the identifier: The path to the
entitity file (use your own, if you don't like it :-) and ".../=
/EN" ->
"...//EN//XML". The last change is what made everything wo=
rk! If you omit
the "OVERRIDE" directive from above, you must change paths=
 in the file
dbcentx.mod too.
 
I have also edited docbook.cat for the SGML version of DocBook: I
commented out the DTDDECL directive, since jade doesn't supportit,
and jade is what I use :-):
 
From
 
&=
nbsp; DTDDECL "-//OASIS//DTD DocBook V4.1//EN" "docbo=
ok.dcl"
 
to
 
&=
nbsp; -- DTDDECL "-//OASIS//DTD DocBook V4.1//EN" "do=
cbook.dcl" --
 
I also edited the paths to the ent=
ities, e.g.:
 
From:
 
&=
nbsp; PUBLIC "ISO 8879:1986//ENTITIES Diacritical Marks//EN&quo=
t;
      =
;   "iso-dia.gml"
 
to
 
&=
nbsp; PUBLIC "ISO 8879:1986//ENTITIES Diacritical Marks//EN&quo=
t;
      =
;   "../../../ISOentities/SGML/ISOdia"
 
Now my DocBook/XML documents conta=
ins the following document type
declaration:
 
  <!DOCTYPE book PUB=
LIC "-//OASIS//DTD DocBook XML V4.1.2//EN"
      =
;            &n=
bsp;     "urn:x-oasis:docbook-xml-v4.1.2&qu=
ot;>
 



When using the XML version of DocBook, I invoke j=
ade like this (line
wrapped, don't try this at home :-)
 
&=
nbsp; jade -t rtf -d /path/to/DSSSL/docbook/print/docbook.dsl
      =
; /path/to/jade/xml.dcl docname
 
Everything works perfectly, even w=
hen I use national (danish) characters
in the input file.
 
I hope you can use this, otherwise give me a call^H^H^H^Hmail!<=
/o:p>
 
Best regards
 
&=
nbsp;   Rune Enggaard Jensen
    r.e.jense=
n@bigfoot.com
 



 



·        
References: 



<=
span
style=3D'mso-list:Ignore'>o&nb=
sp;      
DOCBOOK-APPS: Transforming DocBook/XML with jade on W=
in32=
 



§        
From: Sebastian Bergmann
<sb@sebastian-bergmann.de>



<=
span
style=3D'mso-list:Ignore'>o&nb=
sp;      
RE: DOCBOOK-APPS: Transforming DocBook/XML with jade =
on
Win32 



§        
From: Dave Pawson
<daveP@dpawson.freeserve.co.uk>



<=
span
style=3D'mso-list:Ignore'>o&nb=
sp;      
Re: DOCBOOK-APPS: Transforming DocBook/XML with jade on Win32=
 



§        
From: Norman Walsh
<ndw@nwalsh.com>



 







Appendix B



Automated Configuration Script to Implement Appendix A



 



"""        =
            &nb=
sp;            =
            &nb=
sp;            =
            &nb=
sp;  




File:        
Unpack.py    =
        =
            &nb=
sp;            =
            &nb=
sp;   



        =
            &nb=
sp;            =
            &nb=
sp;            =
            &nb=
sp;   




Purpose:      Unpack all
essential DocBook related zip files into a 



        =
     
working hierarchy.



        =
            &nb=
sp;      
        =
            &nb=
sp;            =
            &nb=
sp;            =
            &nb=
sp;            =
            &nb=
sp;            =
            



Author(s):    Andrew Jonathan Fine        =
            &nb=
sp;            =
       




        =
            &nb=
sp;            =
            &nb=
sp;            
        =
         



DESIGN NOTES: The script implements=
 the
XML portion of suggestions from 



        =
     
OASIS archives application note:



 



        =
     
http://lists.oasis-open.org/archives/docbook-apps/200011/msg00183.ht=
ml



 



        =
     
submitted to OASIS by Rune Enggaard Jensen,



        =
     
<r.e.jensen@bigfoot.com>



        =
     
on Wed 29 Nov 2000 16:33:37 +0100



 



USAGE:        Place
archives adjacent to this script.  The
script will extract all the 



        =
     
information which it needs from archives, 



        =
     
creating a directory, “SGML”, containing a working DocBo=
ok
configuration.  



        =
     
This directory can be copied where needed.



 



"""



 



###



#&=
nbsp;
Standard Python Libraries used by this script:



###



from   strop  import find, replace



from   shutil import copyfile



import sys, os, zipfile<=
/span>



 



 



###



# Specify files to unpack:



###



JADE_BINARY_DISTRIBUTION   =3D 'jadew1_2_1'<=
/span>



DOCBOOK_DSSSL_DISTRIBUTION =3D
'docbook-dsssl-1.73'



DOCBOOK_XML_DISTRIBUTION   =3D 'docbkx412'



DOCBOOK_SGML_DISTRIBUTION  =3D 'docbk41'



ISOENTITIES_DISTRIBUTION   =3D 'ISOEnts'



 



 



###



# Constants:



###



def _dots3 (x): return x [0] + '.' +
x[1] + '.' + x[2]



def _dots2 (x): return x [0] + '.' +
x[1]



 



XML_VERSION_ID        =
    
=3D DOCBOOK_XML_DISTRIBUTION  =
[
-3 : ]



XML_VERSION_NAME        =
  
=3D _dots3  
(XML_VERSION_ID)



 



SGML_VERSION_ID        =
   
=3D DOCBOOK_SGML_DISTRIBUTION [ -2 : ]



SGML_VERSION_N=
AME        =
 
=3D _dots2  
(SGML_VERSION_ID)



 



DO_NOT_COPY        =
       
=3D ('', '')



 



ZIP_EXTN        =
          
=3D '.zip'



PROJECT             =
       =3D
'c:\\usr\packages'  # must alr=
eady
exist!  adjust as needed<=
/o:p>



SGML        =
            &nb=
sp; 
=3D '\\sgml\\'



argv        =
            &nb=
sp; 
=3D sys.argv



args        =
            &nb=
sp; 
=3D len (argv)



WHERE        =
            &nb=
sp;
=3D ((args > 1) and argv [1].strip()) or 'Eggs'=




 



SGML_DESTINATION        =
  
=3D PROJECT +  WHERE + =
SGML



DOCBOOK_DSSSL_DESTINATION  =3D SGML_DESTINATION + 'DSSSL\\Doc=
book'







DOCBOOK_XML_DESTINATION    =3D SGML_DESTINATION +
'DocBook\\XML\\'  + XML_VERSIO=
N_NAME



DOCBOOK_SGML_DESTINATION   =3D SGML_DESTINATION +
'DocBook\\SGML\\' + SGML_VERSION_NAME



ISO_XML_DESTINATION        =3D
SGML_DESTINATION + 'ISOEntities\\XML'



ISO_SGML_DESTINATION       =3D
SGML_DESTINATION + 'ISOEntities\\SGML'



JADE_DESTINATI=
ON        =
  
=3D SGML_DESTINATION + 'Jade'



CATALOG_DESTIN=
ATION        =3D
SGML_DESTINATION + 'catalog.cat'



JADE        =
            &nb=
sp; 
=3D 'jade.exe'



DOCBOOK_CATALO=
G        =
   
=3D 'docbook.cat'



 



catalog        =
           
=3D 'CATALOG "jade/catalog" \n'        =
        
\



        =
            &nb=
sp;     
+ 'CATALOG "DSSSL/docbook/catalog" \n'        \



          =
            &nb=
sp;    +
'CATALOG "Docbook/SGML/'        =
           
\



        =
            &nb=
sp;     
+  SGML_VERSION_NAME        =
            &nb=
sp;    
\



        =
            &nb=
sp;     
+ '/'     =
            &nb=
sp;            =
          
\



        =
            &nb=
sp;     
+  DOCBOOK_CATALOG        =
            &nb=
sp;      
\



        =
            &nb=
sp;     
+ '" \n'   &nbs=
p;            &=
nbsp;           &nbs=
p;        
\



        =
            &nb=
sp;     
+ 'CATALOG "Docbook/XML/'        =
            
\



        =
            &nb=
sp;     
+  XML_VERSION_NAME        =
         
        =
 \



        =
            &nb=
sp;     
+ '/'     =
            &nb=
sp;            =
          
\



        =
            &nb=
sp;     
+  DOCBOOK_CATALOG        =
            &nb=
sp;      
\



        =
            &nb=
sp;     
+ '"'







def content (path):



 



    'content of a file
(non-existent file implies empty content).'



 



    result =3D None



 



    try:=




       
file    =3D o=
pen
(path)



 



    except:



       
return



 



    try:=




       
result  =3D file.read (=
)



 



    except:



        pass



 



    try:=




       
file.close ()



 



    except:



        pass



 



    return result



 



 



 



def emit (path, content, mode =3D
"w"):



 



    """



    create file containing
content, return file's path when successful



    """



 



    result =3D None



 



    if not path:



       
return



 



    if not content:



       
return



 



    try:=




       
file   =3D open (p=
ath,
mode)



 



    except:



       
return result



 



    try:=




       
file.write (content)



       
file.flush ()



       
result =3D path



 



    except:



        pass



 



    return result







###



# Unpack any archive, with optional
filtering action:



#



#        =
   
(name, content) =3D filter (name, content)



###



def unpack (archive, filter =3D Non=
e,
archives_path =3D '', debug =3D 0):



    



    if archives_path:=




       
archive =3D os.path.join  =
; 
(archives_path, archive)



    



    z        =
  
=3D zipfile.ZipFile (archive)



 



    for name in z.namelist=
 ():



       
content =3D z.read (name)



        =
   




        if
filter:



        =
   
(name, content) =3D filter (name, content, archives_path)=




 



        if
name:



 



        =
   
name        =
   
=3D replace (name, '/', '\\')



 



        =
   
if debug:  <=
/span>



        =
  
     print name,
len  (content)



        =
   
else:



        =
       
sys.stdout.write ('.')



 



        =
   
try:



        =
       
os.makedirs (os.path.dirname (name))



 



        =
   
except:



        =
       
pass        #
usually means path already exists



 



        =
 
  emit (name, cont=
ent,
'wb')



 



    z.close  ()



 



 



 



###



# Unpack filter for DocBook DSSSL
Distribution



###



def _docbook_dsssl_filter (name,
content, dummy =3D None): 



    name =3D replace (name,
DOCBOOK_DSSSL_DISTRIBUTION, DOCBOOK_DSSSL_DESTINATION)



    return (name, content)=




 



 



 



###



# Unpack filter for DocBook XML
Distribution



###



def _docbkx_filter       
(name, content, dummy =3D None):



 



 



    ax      =3D 'PUBLIC
"-//OASIS//DTD DocBook XML V' \



        =
   
+ XML_VERSION_NAME        =
            &nb=
sp;
\



        =
   
+ '//EN" "docbookx.dtd"'



 



    ay      =3D 'OVERR=
IDE
YES\n'        =
            &nb=
sp;
\



        =
   
+ 'SYSTEM "urn:x-oasis:docbook-xml-v'   \



        =
   
+ XML_VERSION_NAME        =
            &nb=
sp;
\



        =
   
+ '" "docbookx.dtd"\n\n'        =
       
\



        =
   
+  ax=




 



&n=
bsp;
  b00x   =3D 'Diacritical Marks//EN&q=
uot;'



    b00y   =3D 'Diacritical
Marks//EN//XML"'



 



    b01x   =3D 'Numeric and Special
Graphic//EN"'



    b01y   =3D 'Numeric and Special
Graphic//EN//XML"'



 



    b02x   =3D 'Publishing//EN"'



    b02y   =3D 'Publishing//EN//XML&quo=
t;'



 



    b03x   =3D 'General Technical//EN&q=
uot;'



    b03y   =3D 'General
Technical//EN//XML"'



 



    b04x   =3D 'Added Latin 1//EN"=
' 



    b04y   =3D 'Added Latin 1//EN//XML&=
quot;'



 



    b05x   =3D 'Added Latin 2//EN"=
' 



    b05y   =3D 'Added Latin 2//EN//XML&=
quot;'



    b06x   =3D 'Greek Letters//EN"=
' 



    b06y   =3D
'Greek Letters//EN//XML"'



 



    b07x   =3D 'Monotoniko Greek//EN&qu=
ot;' 



    b07y   =3D 'Monotoniko
Greek//EN//XML"'



 



    b08x   =3D 'Greek Symbols//EN"=
' 



    b08y   =3D 'Greek Symbols//EN//XML&=
quot;'



 



    b09x   =3D 'Alternative Greek
Symbols//EN"' 



    b09y   =3D 'Alternative Greek
Symbols//EN//XML"'



 



    b10x   =3D 'Added Math Symbols: Arr=
ow
Relations//EN"' 



    b10y   =3D 'Added Math Symbols: Arr=
ow
Relations//EN//XML"'



 



    b11x   =3D 'Added Math Symbols: Bin=
ary
Operators//EN"' 



    b11y   =3D 'Added Math Symbols: Bin=
ary
Operators//EN//XML"'



 



    b12x   =3D 'Added Math Symbols:
Delimiters//EN"' 



    b12y   =3D 'Added Math Symbols:
Delimiters//EN//XML"'



 



    b13x   =3D 'Added Math Symbols: Neg=
ated
Relations//EN"' 



    b13y   =3D 'Added Math Symbols: Neg=
ated
Relations//EN//XML"'



 



    b14x   =3D 'Added Math Symbols:
Ordinary//EN"' 



    b14y   =3D 'Added Math Symbols:
Ordinary//EN//XML"'



 



    b15x   =3D 'Added Math Symbols:
Relations//EN"' 



    b15y   =3D 'Added Math Symbols:
Relations//EN//XML"'



 



    b16x   =3D 'Box and Line Drawing//E=
N"'




    b16y   =3D 'Box and Line
Drawing//EN//XML"'



 



    b17x   =3D
'Russian Cyrillic//EN"' 



    b17y   =3D 'Russian
Cyrillic//EN//XML"'



 



    b18x   =3D 'Non-Russian Cyrillic//E=
N"'




    b18y   =3D 'Non-Russian
Cyrillic//EN//XML"'



 



    cx     =3D 'ent/iso-'



    cy     =3D '../../../IS=
OEntities//XML//iso-'



 



    if name =3D=3D DOCBOOK=
_CATALOG:



        for
pair in [(ax,    ay)=
,  (b00x, b00y), (b01x, b01y), <=
/o:p>



        =
            
(b02x, b02y),
(b03x, b03y), (b04x, b04y),



        =
            
(b05x, b05y), (b06x, b06y), (b07x, b07y),



        =
            
(b08x, b08y), (b09x, b09y), (b10x, b10y),



        =
            
(b11x, b11y), (b12x, b12y), (b13x, b13y),



        =
            
(b14x, b14y), (b15x, b15y), (b16x, b16y),



        =
            
(b17x, b17y), (b18x, b18y), (cx,   cy)]:   



          =
  content
=3D replace (content, pair [0], pair [1])



 



    if
find (name, '\\ent')  >=3D =
0:



       
return DO_NOT_COPY



 



    if find (name, '.ent')=
   < 0:



        name
=3D os.path.join (DOCBOOK_XML_DESTINATION, name)



    else:



        name
=3D os.path.join (ISO_XML_DESTINATION, os.path.basename (name))<=
/span>



 



    return (name, content)=








###



# Unpack filter for DocBook SGML
Distribution



###



def _docbk_filter       
(name, content, dummy =3D None):



 



    ax     =3D 'DTDDECL
"-//OASIS//DTD DocBook V4.1//EN" "docbook.dcl"'



    ay   =
 
=3D '-- ' + ax + ' --'



 



    cx     =3D '"iso-'=




    cy     =3D
'"../../../ISOEntities//SGML//iso'



 



    if name =3D=3D DOCBOOK=
_CATALOG:



        for
pair in [(ax, ay), (cx, cy)]:  




        =
   
content =3D re=
place
(content, pair [0], pair [1])



 



    name
=3D os.path.join (DOCBOOK_SGML_DESTINATION, name)



    return (name, content)=




 



 



###



# Unpack filter for Jade distributi=
on



###



def _jade_filter        =
 
(name, info, archives_path):



    name =3D os.path.join<=
span
style=3D'mso-spacerun:yes'>  (JADE_DESTINATION, name)



 



    if find      (name, JAD=
E)
>=3D 0:



       
path        =
          
=3D 'Patches\\' + JADE



        if
archives_path:  



        =
   
path =3D os.path.join (archives_path, path) 



 



       
copyfile (path, name)



       
return   DO_NOT_CO=
PY



 



    return (name, info)



 



 



 



###



# Unpack filter for ISO Entities



###



def _iso_filter        =
  
(name, info, dummy =3D None): =
 



    name =3D os.path.join<=
span
style=3D'mso-spacerun:yes'>  (ISO_SGML_DESTINATION, name + '.gm=
l')



    return (name, info)



 



 



 



###



# Lay down paths for file tree=




###



def _buildtree ():



    for x in [
DOCBOOK_XML_DESTINATION,   \



        =
      
DOCBOOK_SGML_D=
ESTINATION,  \



        =
      
DOCBOOK_DSSSL_DESTINATION, \



        =
      
ISO_SGML_DESTINATION,      \



        =
      
JADE_DESTINATION ]:



 



        try:



        =
   
os.makedirs (x)



 



       
except:



        =
   
pass        =
  
# usually means path already exists







###



# Procedural interface:<=
/span>



###



def Unpack (archives_path =3D '', d=
ebug =3D
0):



    print     "Unpacking
DocBook Tools",



 



    buildtree ()



    emit      
(CATALOG_DESTINATION,        
catalog)



 



    for pair in [ (DOCBOOK=
_XML_DISTRIBUTION   + ZIP_EXTN,   \



        =
          
_docbkx_filter),        =
            &nb=
sp;   
\



        =
            &nb=
sp;            =
            &nb=
sp;            =
 
\



        =
         
(DOCBOOK_SGML_DISTRIBUTION  +
ZIP_EXTN,   \<=
/span>



        =
          
_docbk_filter),    <=
/span>        =
            &nb=
sp; \



        =
            &nb=
sp;            =
            &nb=
sp;            =
 
\



        =
         
(DOCBOOK_DSSSL_DISTRIBUTION + ZIP_EXTN,   \



        =
          
_docbook_dsssl_filter),        =
         
\



        =
            &nb=
sp;            =
            
        =
      \



        =
         
(JADE_BINARY_DISTRIBUTION &nbs=
p;
+ ZIP_EXTN,   \



        =
          
_jade_filter),        =
            &nb=
sp;     
\



        =
            &nb=
sp;            =
            &nb=
sp;            =
 
\



        =
         
(ISOENTITIES_DISTRIBUTION &nbs=
p;
+ ZIP_EXTN,   \



        =
          
_iso_filter),        =
            &nb=
sp;      
\



        =
       
]:



 



       
unpack (pair [0], pair [1], archives_path, debug)<=
/p>


 



 



###



# Command-line interface:



###



if __name__ =3D=3D '__main__':=




    Unpack (debug =3D 1)



 



#   End of 'Unpack.py'   #

Appendix C

Local modifications to OpenJade

 

The vario= us teams wanted documents containing features that DocBook and OpenJade did not ordinarly provide but were routinely expected in Word documents.  The features included:

&nbs= p;

·        Identifiers that would display as entries in an index.

·        Figures and tables whose captions w= ould display as entries in a table of contents.

&nbs= p;

These spe= cial features were normally implemented as field codes in the Microsoft Word obj= ect model.  However, OpenJade did = not implement these field codes in its Rich Text Format backend generator.  

&nbs= p;

The follo= wing patch modifies the RTF backend to allow a DocBook stylesheet to arbitrarily= emit special characters that are essential components in RTF implementations of = Word field codes.

&nbs= p;

In file RtfFOTBuilder.cxx

In method= void RtfFOTBuilder::characters ()=

&nbs= p;

CHANGE FROM

    case '\0':

      break;

    case ';':=

    case ',':=

      if (math= Level_ && *s =3D=3D eqArgSep_ && mathSpecial_ =3D=3D mathNormal)=

 

CHANGE INTO

      break;

 

///////// DANGER!!!  MIS EN GUARDE!!!   ACHCTUNG!!!  PELIGROSO!!! //////////

    case 0xfffd:

        os() << '{';    // Left  brace for user-embedded Word 97 = Field codes

        break;

 

      case 0xff= fe:

        os() << '}';    // Right brace for user-embedded Word 97 Field codes

        break;

 

      case 0xff= ff:

        os() << '\\';   // Backslash for user-embed= ded Word 97 Field codes

        ///////// DANGER!!!  MIS EN GUARDE!!!   ACHCTUNG!!!  PELIGROSO!!! //////////

    case ';':=

    case ',':=

      if (math= Level_ && *s =3D=3D eqArgSep_ && mathSpecial_ =3D=3D mathNormal)=

 

Normally, the wide characters for left curly brace {, ri= ght curly brace }, and backstroke \, are emitted with a backstroke preceding th= em ( \{, \}, and \\ respectively) so that RTF sees them as escaped literal text rather than as part of an RTF expression.

&nbs= p;

To emit arbitrary field codes at will, I needed to define wide characters that map = to the unescaped versions.  Wide character 0xFFFD maps to {, wide character 0xFFFE maps to }, and wide chara= cter 0xFFFF maps to \ when OpenJade is recompiled with this patch.

&nbs= p;

This patc= h is a gateway that now can allow someone who does not use extreme care, to caus= e great mayhem within an RTF document and hence to Word when the document is read.<= o:p>

&nbs= p;

Appendix D and E together form an excellent example of how to safely interface this new feature to DocBook.  Appendix = D adds some new high level terms in DocBook that are used by the Python-generated DocBook SGML.  Appendix E impl= ements several modifications to the DocBook <para> production to add special effects such as indexable text.


Appendix D

Local DocBook stylesheet definitions (Local.dtd)

 

 

<!--

;; File:      Local.dtd

;;

;; Purpose:   DocBook DTD declaration customization layer specific

;;        =     to local documentation requirements

;;

;; Author(s): Andrew Jonathan Fine

;;-->        =          

 

<!--

;;

;; DESIGN NOTES:

;;

;; The open source text processor, Jade, requires an input text

;; file, a DTD file, and a DSL file in order to operate properly.

;;

;; The chain of dependencies you have to consider when processing

;; any input text file goes like this:

;;

;; text file =3D=3D=3D>  DTD f= ile =3D=3D=3D> DSL file

;;

;; - The purpose of the text file is obvious.  

;;

;; - The purpose of the DTD file is to declare any and all SGML tags

;;   (enclosed by angle brackets) used b= y the text file as having

;;   some kind of purpose. 

;;

;; - The purpose of the DSL file is to define the specific behavior=

;;   (using the DSSSL language) o= f the tags declared by the DTD.

;;

;;

;; The text file points to the DTD file like this:

;;       (yes, foward / is required by Jade, this is not a bug!)

;;

;;       !DOC= TYPE section SYSTEM "c:/Local.dtd"

;;

;;        =       Which reads, literally, "the definition of the tag

;;        =       'section' is found in the DTD file whose system-level

;;        =       location is "c:/Local.dtd"

;;

;;        =       All the other tags are "enclosed" by the section tag,=

;;        =       which is why no other file references are required.

;;

;;

;; The DTD file refers to element definitions located in any DSL file

;; which Jade can find.  That DSL= file is specified on the Jade command

;; line using the '-d' option.  

;;

;;

;; Not all elements need to be declared in the top-level DTD file, and

;; not all elements need to be defined in the top-level DSL file.  This

;; allows for site-specific modifications of text-processing behavior.

;;

;;

;; (the file, xml.dcl, is a Jade standard file used by Jade to define

;; the syntax of XML, which is the meta-language the input text file is

;; written in).

;;

;;

;; Finally, the '-t rtf' option in Jade allows for the generation of

;; Microsoft Word Rich Text Format, as the kind of output we want. =

;;

;;

;; Tying this all together is the following command-line:

;; =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D

;;

;; jade -t rtf -d c:/Local.dsl c:/usr/packages/sgml/jade/xml.dcl input.txt

;;

;;

;; To understand further on writing custom DSSSL, please refer to

;; Section 4.3.8, 'Customizing the Stylesheets', from 'DocBook, The

;; Definitive Guide', published by O'Reilly & Associates, (c) 2001.

;;-->

<!--

   Declare a local instance of general character class

   whose name is textfield.

  -->

<!ENTITY % local.gen.char.class  "|textfield"> &nb= sp;

 

        =   

        =   

<!--

   Read the existing declaratio= ns for DocBook from Jade's general catalog.

  -->        =             &nb= sp;         

<!ENTITY % DocBookDTD PUBLIC "-//OASIS//DTD DocBook XML V4.1.2//EN" "urn:x-oasis:docbook-xml-v4.1.2">

%DocBookDTD;

 

 

        =    

<!--

   Declare the behavior of the element textfield as based on a

   standard behavior offered by DocBook (refer to The Definitive

   Guide appendices for details= on how %refinline.char.mix works)

  -->        =    

<!ELEMENT textfield - - ((%refinline.char.mix;)+)&nb= sp; >

 

 

    

<!--

   Declare that the attributes = which textfield can use are the

   common ones which most eleme= nts in DocBook can use.

  -->

<!ATTLIST textfield        =             &nb= sp;          

        %common.attrib;

> 

 

 

      

<!--  End of "Local.dtd"   -->


Appendix E

Local DocBook stylesheet implementations (Local.dsl)

&nbs= p;

<!DOCTYPE style-sheet PUBLIC "-//James Clark//DTD DSSSL Style Sheet//EN" [<= o:p>

<!ENTITY dbstyle PUBLIC "-//Norman Walsh//DOCUMENT DocBook Print

Stylesheet//EN" CDATA DSSSL>

]>

 

 

<style-sheet>

<style-specification use=3D"docbook">

<style-specification-body>

 

 

;;;;

;; File:      Local.dsl

;;

;; Purpose:   DocBook DSL style-sheet customization layer specific

;;        =     to local documentation requirements

;;

;; Author(s): Andrew Jonathan Fine

;;;;

 

 

;;;;

;;

;; DESIGN NOTES:

;;

;; The open source text processor, Jade, requires an input text

;; file, a DTD file, and a DSL file in order to operate properly.

;;

;; The chain of dependencies you have to consider when processing

;; any input text file goes like this:

;;

;; text file --->  DTD file --= -> DSL file

;;

;; - The purpose of the text file is obvious.  

;;

;; - The purpose of the DTD file is to declare any and all XML tags

;;   (enclosed by angle brackets)= used by the text file as having

;;   some kind of purpose. 

;;

;; - The purpose of the DSL file is to define the specific behavior=

;;   (using the DSSSL language) o= f the tags declared by the DTD.

;;

;;

;; The text file points to the DTD file like this:

;;       (yes, foward / is required by Jade, this is not a bug!)

;;

;;       !DOC= TYPE section SYSTEM "c:/Local.dtd"

;;

;;        =       Which reads, literally, "the definition of the tag

;;        =       'section' is found in the DTD file whose system-level

;;        =       location is "c:/Local.dtd"

;;

;;             =  All the other tags are "enclosed" by the section tag,

;;        =       which is why no other file references are required.

;;

;;

;; The DTD file refers to element definitions located in any DSL file

;; which Jade can find.  That DSL= file is specified on the Jade command

;; line using the '-d' option.  

;;

;;

;; Not all elements need to be declared in the top-level DTD file, and

;; not all elements need to be defined in the top-level DSL file.  This

;; allows for site-specific modifications of text-processing behavior.

;;

;;

;; (the file, xml.dcl, is a Jade standard file used by Jade to define

;; the syntax of XML, which is the meta-language the input text file is

;; written in).

;;

;;

;; Finally, the '-t rtf' option in Jade allows for the generation of

;; Microsoft Word Rich Text Format, as the kind of output we want. =

;;

;;

;;

;; Tying this all together is the following command-line:

;; ------------------------------------------------------

;; jade -t rtf -d c:/Local.dsl c:/usr/packages/sgml/jade/xml.dcl input.txt

;;

;;

;;

;; To understand further on writing custom DSSSL, please refer to

;; Section 4.3.8, 'Customizing the Stylesheets', from 'DocBook, The

;; Definitive Guide', published by O'Reilly & Associates, (c) 2001.

;;;;

 

 

 

;;;

;  Parameters to DSSSL rulesets<= /o:p>

;;;

(define %title-font-family%   "Times New Roman")

(define %body-font-family%    "Times New Roman")

(define %mono-font-family%    "Courier New")

(define %admon-font-family%   "Arial")

(define %dingbat-font-family% "WingDings")

(define %body-start-indent%    0pt)        =   ; rely on Word template for indentation

 

 

 

;;;

;  Parameters to Jade RTF backend

;;;

(declare-characteristic

       prese= rve-sdata?

       "= ;UNREGISTERED::James Clark//Characteristic::preserve-sdata?&q= uot;

       #t

)

 

 

 

;;;;

;; A DSSSL function to write a custom XML tag, 'textfield', along

;; the general style of a paragraph - with specific appearance.

;;

;; Note:  The custom tag, 'textfi= eld', in defined in Local.dtd

;;;;

(element textfield

  (let ((effect (attribute-string (normalize "condition"))))

    (if (not effect)<= /o:p>

        (make paragraph)

        (case effect

;         (("6")        =   (make paragraph font-size:     6pt))=

;         (("8")        =   (make paragraph font-size:     8pt))=

        =   (("page-break") (make paragraph break-before: 'page))=

        =   (else        =    (make paragraph))))))

 

 

 

;;;;

;; A DSSSL override function to an existing element,

;; to center images in the print output

;;

;; See DocBook StyleSheet FAQ, Section 1, from

;; http:/www.miwie.org/docbook-dsssl-faq.html#CENTERIMAGES

;;;;

(element imagedata

  (if (have-ancestor? (normalize "mediaobject"))

    ($img$ (current-node) #t)        =         

    ($img$ (current-node) #f)))  

   

 

 

;;;;

;; make:  \code content

;;;;

(define  (rtf-escape-code   code content)

  (let ((stroke  (literal "\U-FFFF"))

        (pad     (literal " ")))

    (make sequence stroke = code pad content)))

 

 

 

;;;;

;; make: { content }

;;;;

(define  (rtf-brace-content content)

  (let ((lbrace  (literal "\U-FFFD"))

        (rbrace  (literal "\U-FFFE")))

    (make sequence lbrace content rbrace)))

   

 

;;;;

;; make: \code { content }

;;;;

(define  (rtf-escape-code-brace-content code content)

&n= bsp; (rtf-escape-code

   code

   (rtf-brace-content content))= )

  

 

;;;;

;; make: { \code { content } }=

;;;;

(define  (simple-field-code code content)

&n= bsp; (rtf-brace-content

    (rtf-escape-code-brace-content code content)))

   

 

;;;;

;; make: code \modifier<= /span>

;;;;

(define (rtf-code-modifier  content modifier)

&n= bsp; (let ((stroke  (literal "\U-FFFF"))

        (pad     (literal " ")))

    (make sequence content stroke modifier)))

 

 

;;;;

;; make: code \modifier parameter

;;;;

(define (rtf-code-parameter content modifier parameter)

  (let ((stroke  (literal "\U-FFFF"))

        (pad     (literal " ")))

    (make sequence content stroke modifier pad parameter)))

   

 

 

;;;;

;; DSSSL function to create index field codes

;;;;

(define (index-field-code content effect)

  (let  ((index   (literal  "xe" ))

         (italic  (literal  "ixe"))

         (bold    (literal  "bxe&quo= t;)))

           =      

    (simple-field-code

     index=

    

     (case effect

       

      (("italic")         (rtf-code-modifier content italic))

      (("bold")        =    (rtf-code-modifier content bold))

     

      (("bold-italic")

       (rtf-code-modifier (rtf-code-modifier content italic) bold))

        

      (else        =         content)))))

      

 

;;;;

;; DSSSL function to create index field effect

;;;;

(define (index-field content effect)

  (index-field-code (rtf-brace-conte= nt content) effect))

 

 

;;;;

;; DSSSL function to create index field for specific index

;;;;

(define (select-index-field content effect selector)

  (index-field

    (rtf-code-parameter co= ntent (literal "xef") (literal selector))

    effect))

;;;;

;; DSSSL function to create RTF/Word hidden text field

;;;;

(define  (hidden-text   content)

  (let ((lbrace  (literal "\U-FFFD"))

        (rbrace  (literal "\U-FFFE"))

        (stroke  (literal "\U-FFFF"))

        (pad     (literal " "))

        (hide    (lit= eral "v")))

    (make sequence lbrace = stroke hide pad content rbrace))) 

   

        =  

 

;;;;

;; DSSSL function to create RTF/Word field codes

;;;;

(define  (extend-field-code code content)

  (let ((lbrace  (literal "\U-FFFD"))

        (rbrace  (literal "\U-FFFE"))

        (stroke  (literal "\U-FFFF"))

        (qmark   (literal "\""))

        (field   (literal "field"))

        (fldinst (literal "fldinst"))

        (star    (lit= eral "*"))

        (pad     (literal " ")))

    (make sequence

        =   lbrace stroke  field

          lbrace stroke  star

        =          stroke  fldinst

        =   lbrace code    pad

        =   qmark  content qmark

        =   rbrace

        =   rbrace

        =   rbrace)))

        =          

 

;;;;

;; DSSSL override to apply special effects to a phrase if desired.<= /span>

;;;;

(element phrase

  (let ((role    (attribute-string (nor= malize "role")))

        (effect  (attribute-str= ing (normalize "condition")))

        (xref    (attribute-string (normalize "xreflabel"))))

 

    (if (not role)

        ($charseq$)

        

        (case role

       

        =   (("xe")    (index-field        =    ($charseq$) effect))

        =  

        =   (("xef")   (select-index-field  &nbs= p; ($charseq$) effect xref))

        =  

        =   (("xei")   (index-field        =    (hidden-text ($charseq$)) effect))

        =  

        =   (("xeif")  (select-index-field  &nbs= p; (hidden-text ($charseq$)) effect xref))        =             &nb= sp;            =     

 

        =   (("seq")   (extend-field-code     (literal role) <= o:p>

        =             &nb= sp;            =            ($charseq$)))

        =             &nb= sp;            =           

        =   (("toc")   (extend-field-code     (literal "t= oc   \\c")=

        =             &nb= sp;            =            ($charseq$)))

        =             &nb= sp;            =           

        =   (("index") (extend-field-code     (literal "I= NDEX \\c")

        =             &nb= sp;            =            ($charseq$))) 

        =             &nb= sp;            =       

        =   (("doc")   (extend-field-code     (literal "I= NCLUDEPICTURE")

        =             &nb= sp;            =            ($charseq$)))

        =             &nb= sp;            =            

        =   (("section-break")

        =              (rtf-escape-code       (lit= eral "sect")

        =             &nb= sp;            =            ($charseq$)  ))

        =       

        =   (("page-break")

        =            =   (rtf-escape-code       (lit= eral "page")

        =             &nb= sp;            =            ($charseq$)  ))

        =          

        =   (else      ($charseq$))))))

   

 

 

;;;;

;; A DSSSL override for treatment of sections and section titles.

;;

;; (The original text was copied from DocBook dbsect.dsl

;;;;

(define ($section-title$)

  (let* ((sect (current-node))<= /o:p>

        (info (info-element))

        (exp-children (if (node-list-empty?= info)

        &= nbsp;               (empty-node-list)=

        &= nbsp;               (expand-children (children i= nfo)

        &= nbsp;           &nbs= p;             =     (list (normalize "bookbiblio")

        &= nbsp;           &nbs= p;            &= nbsp;         (normalize "bibliomisc")=

        &= nbsp;           &nbs= p;            &= nbsp;         (normalize "biblioset"))= )))

        (parent-titles (select-elements (ch= ildren sect) (normalize "title")))

        (info-titles   (select-elements exp-children (normalize "title")))

        (titles        (if (node-list-empty? parent-titles= )

        &= nbsp;                info-titles=

        &= nbsp;                parent-titles))

        (subtitles     (select-elements exp-children (normalize "subtitle")))

        (renderas (inherited-attribute-stri= ng (normalize "renderas") sect))

        (hlevel        =             &nb= sp;     ;; the apparent section level;

         (if renderas        =             ;; if not real section level,

             (string->number        =      ;;   then get the apparent level

              (sub= string renderas 4 5))  ;;   from "renderas",

             (SECTLEVEL)))        =        ;; else use the real l= evel

        (hs (HSIZE (- 4 hlevel))))

    (make sequence

      (make para= graph

       font-= family-name: %title-font-family%

       font-= weight:  'medium

       font-= posture: 'upright

       font-= size:       11.5= pt

       line-= spacing: (* hs %line-spacing-factor%)

       space= -before: (* hs %head-before-factor%)

       space= -after: (if (node-list-empty? subtitles)

        &= nbsp;             (* hs %head-after-factor%)

        &= nbsp;             0pt)

       start= -indent: (if (or (>=3D hlevel 3)

        &= nbsp;                  (member (g= i) (list (normalize "refsynopsisdiv")

        &= nbsp;           &nbs= p;            &= nbsp;        (normalize "refsect1")

        &= nbsp;                    &= nbsp;             (normalize "refsect2")

        &= nbsp;           &nbs= p;            &= nbsp;        (normalize "refsect3"))))=

        &= nbsp;              %body-start-indent%

        &= nbsp;              0pt)

       first= -line-start-indent: 0pt

       quadd= ing: %section-title-quadding%

       keep-= with-next?: #t

       headi= ng-level: (if %generate-heading-level% (+ hlevel 0) 0)

       ;; Si= mpleSects are never AUTO numbered...they aren't hierarchical

       (if (string=3D? (element-label (current-node)) "")<= /p>

           (empty-sosofo)

           (literal (element-label (current-node))

        &= nbsp;          (gentext-label-title-sep (gi sect))))

       (elem= ent-title-sosofo (current-node)))

      (with-mode section-title-mode

       (proc= ess-node-list subtitles))

      ($section-= info$ info))))

 

 

 

;;;;

;; After our own custom modifications, read in the general DocBook<= /span>

;; DSL definitions, for the general DocBook style.

;;;;

 

</style-specification-body>

</style-specification>

<external-specification id=3D"docbook" document=3D"dbstyle">

</style-sheet>

Appendix F

Example DocBook Class Library

 

The code listed in this appendix contains a working portion of an DocBook SGML gener= ator as implemented in Python, enough to that an one-to-one mapping from DocBook into Python callable classes can be designed without great difficulty.=

&nbs= p;

These scr= ipts are meant to be placed in the same directory.  DocBook SGML applications should i= mport and inherit from class DocBook and its members.

 

 

# ------- Option.py start --------#=

 

class Option:

 

    'an SGML tag option'

 

    def __init__ (self, va= lue):

 

        self.title =3D self.__class__.__name__.lower ()

        self.value =3D value

 

 

    def __str__  (self):

 

        return "%s=3D'%s'" % (self.title, str (self.value)) <= /o:p>

 

# ------- Option.py end --------#

 

 

 

# ------- List.py start --------#

 

class List:

 

    'SGML tag options'

   

    OPTIONS =3D []

    sp      =3D ' '

   

    def __init__ (self, *a= rgs):

 

        self.data =3D [ x (a) for (x, a) in zip (self.OPTIONS, *args) ]=

 

 

    def __str__  (self):

 

        return self.sp.join (map (str, self.data))

 

# ------- List.py end --------#

 

#   --- Lists.py  start --  #

 

from List    import List =

from Options import Options

 

class Lists:

 

    class ColSpec       (Lis= t):

 

        OPTIONS =3D [ Options.ColName, Options.ColWidth, Options.Align ]

     

 

    class Phrase        (List):

   

        OPTIONS =3D [ Options.Role,    Options.Condition ]

 

 

    class TextField     (List):

 

        OPTIONS =3D [ Options.Condition ]

 

 

    class TGroup        (List):

 

        OPTIONS =3D [ Options.Cols, &n= bsp; Options.ColSep, Options.RowSep, Options.Align ]

 

 

    class InformalTable (L= ist):

 

        OPTIONS =3D [ Options.Frame ]

 

 

#   --- Lists.py  end --  #

 

# ------- Options.py start --------#

 

from Option import Option

 

class Options:

 

    class ColName   (Option):<= /p>

      'name of c= olumn'

   

    class ColWidth  (Option):

      'width of column'

   

    class Align     (Option):

      'text alig= nment'

   

    class Role      (Option):<= o:p>

      'escape parameter'

   

    class Condition (Optio= n):

      'escape parameter'

   

    class Cols      (Option):<= o:p>

      'tgroup nu= mber of columns'

   

    class ColSep    (Option):

      'tgroup co= lumn seperation'

   

    class RowSep    (Option):

      'tgroup row seperaton'

   

    class Frame     (Option):

      'informal = table option'

 

# ------- Options.py end --------#

 

#  --- Rule.py start -- #<= /span>

 

from List import List

 

class Rule:

 

    'DocBook production ru= le'

 

    cr           =3D '= \n'

    sp        =    =3D ' '

    nil        =   =3D ''

    LEFT         =3D '<'

    RIGHT        =3D '>'

    STROKE       =3D = '/'

    DOT        =   =3D '.'

    BEGIN        =3D LEFT

    END        =   =3D LEFT + STROKE

    CONTENT      =3D nil

    OPTIONS      =3D List

    IS_TOP_LEVEL =3D False

    IS_INLINE    =3D False

    TITLE        =3D nil

 

   

    def __init__  (self, *args):

 

        self.options =3D self.OPTIONS (args)

        self.title   =3D self._title  ()

        self.data    = =3D []

 

 

    def _title    (self):

 

        result       =3D self.TITLE or self.__class__.__name__

        result       =3D = str (result).lower ()

        return result.split (self.DOT)[-1]

 

 

    def start     (self):

 

        options    &nbs= p; =3D str (self.options).strip ()

        if options:  options =3D self.sp + options

        return self.BEGIN + self.title + options + self.RIGHT

 

   

    def end       (sel= f):

 

        return self.END   + self.title + self.RIGHT

 

 

    def enlist    (self, data):

 

        if not data:  return <= /span>

        if type   (data) !=3D type ([]):  return [ data ]

        return        data

 

 

    def content   (self):

 

        data =3D self.enlist (self.data)

        if data in [ None, [None]]:  retu= rn self.nil

        return self.cr.join (map (self.render, data))

 

 

    def render    (self, content =3D Non= e):

 

        content =3D content or self.CONTENT or self.content

        if not content:         return self.nil

        if callable (content):  return co= ntent ()

        return content

 

 

    def __call__  (self, content =3D None):

 

        result    =3D self.render (content)

        if self.IS_TOP_LEVEL:  return res= ult

        separator =3D [ self.cr, self.nil ] [self.IS_INLINE or (not result)]=         =  

        return separator.join ([self.start (), result, self.end ()])

 

 

#  --- Rule.py end -- #

#   --- Rules.py  start --  #

 

from Lists import Lists

from Rule  import Rule

 

 

class ColSpec     (Ru= le):

 

    'DocBook column specification'

 

    OPTIONS =3D Lists.ColS= pec

 

 

class Phrase      (Rule):

 

    'DocBook text phrase'<= o:p>

 

    OPTIONS =3D Lists.Phra= se

    IS_INLINE =3D True

 

 

class Emphasis    (Phrase)= :

 

    'DocBook emphasized te= xt'

 

 

class Textfield   (Phrase):

 

    'Local DocBook special purpose text'

 

 

class Entry       (Rule):

 

    'column entry in row of table'

 

    IS_INLINE =3D True

 

 

class Row         (Rule):

 

    'row of a DocBook tabl= e'

 

    def __init__   (self, entries):<= /span>

 

        Rule.__init__ (self)

        self.data =3D entries

 

 

class THeadEntry     &n= bsp; (Entry):

 

    'DocBook table heading label'

 

    TITLE        =             =3D str (Entry)

 

    def __init__       (sel= f, name):

 

        self.CONTENT         =3D Emphasis ('bold')

        self.CONTENT.CONTENT =3D name

        Entry.__init__ (self)        =   

 

 

class THeadRow         (Row):

 

    'DocBook table heading= row of heading labels'

 

    TITLE                =     =3D str (Row)

 

    def __init__       (sel= f, names):

 

        Row.__init__   (se= lf, map  (THeadEntry, names))=

 

 

class THead        =     (Rule):

 

    'DocBook table heading= '

 

    def __init__       (sel= f, names):

 

        self.CONTENT        =3D THeadRow (names)

        Rule.__init__  (self)

class TBody        =     (Rule):

 

    'DocBook table body'

 

    def __init__       (sel= f, items):

 

        Rule.__init__  (self)

        self.data =3D items

 

 

class TGroup        =    (Rule):

 

    'DocBook table group'<= o:p>

 

    OPTIONS  =3D Lists.TGroup=

    THEAD    =3D THead

    TBODY    =3D TBody

    SHAPE    =3D []

    COLSPECS =3D []

 

    def __init__       (sel= f, rows):

 

        Rule.__init__  (self, *self.SHAPE)

 

        self.data =3D [ self.cr.join (x () for x in self.COLSPECS),

        =             &nb= sp; self.THEAD ([x.options.data [0].value for x in self.COLSPECS]),=

        =             &nb= sp; self.TBODY (rows)

        =             ]

 

 

 

    def __call__       (sel= f, content =3D None):

 

        return Rule.__call__  (= self, content)

 

  

 

class InformalTable    (Ru= le):

 

    'DocBook informal tabl= e'

 

    OPTIONS  =3D Lists.InformalTable=

    TGROUP   =3D TGroup=

 

    def __init__       (sel= f, *args):

 

        args =3D args or ['all']

        Rule.__init__  (self, *= args)

 

 

    def __call__       (sel= f, rows):

 

        self.data =3D self.TGROUP (rows)

        return Rule.__call__  (= self)

 

 

#   --- Rules.py  end --  #

#  --- DocBook.py start -- #

 

from Rule    import Rule<= o:p>

 

from Rules   import ColSpec       as _ColSpec

from Rules   import Phrase        as _Phrase

from Rules   import Emphasis      as _Emphas= is

from Rules   import Textfield<= span style=3D'mso-spacerun:yes'>     as _Textfield

from Rules   import Entry         as _Entry

from Rules   import Row        =    as _Row

from Rules   import THeadEntry=     as _THeadEntry

from Rules   import THeadRow      as _THeadR= ow

from Rules   import THead         as _THead

from Rules   import TBody         as _TBody

from Rules   import TGroup        as _TGroup

from Rules   import InformalTa= ble as _InformalTable

 

from Options import Options      as _Option= s

 

 

class DocBook (Rule):

 

    Options        =    =3D _Options

    HEADER        =     =3D '<!DOCTYPE %s SYSTEM "C:\Local.dtd">'=

    SECTION        =    =3D 'section'

    IS_TOP_LEVEL      =3D True

 

    class Rules:

 

        ColSpec       =3D<= span style=3D'mso-spacerun:yes'>  _ColSpec

        Phrase        =3D  _Phrase

        Emphasis      =3D  _Emphasis

        Textfield     =3D  _Textfield

        Entry         =3D  _Entry<= /span>

        Row        =    =3D  _Row

        THeadEntry    =3D  _THeadEntry

        THeadRow    &nb= sp; =3D  _THeadRow

        THead         =3D  _THead<= /span>

        TBody=          =3D  _TBody<= /span>

        TGroup        =3D  _TGroup=

        InformalTable =3D  _InformalTable

 

   

    def __init__ (self, da= ta):

 

        Rule.__init__ (self)

        self.data =3D data

 

 

    def _section (self):

 

        result       =3D self.SECTION

        result       =3D = str (result).lower ()

        return result.split (self.DOT)[-1]

 

 

    def header   (self):

 

        return self.HEADER % self._section ()

 

 

    def __call__ (self):

 

        return self.header () + self.cr + Rule.__call__ (self)

 

  

#  --- DocBook.py start -- #

 

 

------=_NextPart_01C52880.DC295B70 Content-Location: file:///C:/C9344E54/Paper14_files/image006.gif Content-Transfer-Encoding: base64 Content-Type: image/gif R0lGODlhQAIhAXcAMSH+GlNvZnR3YXJlOiBNaWNyb3NvZnQgT2ZmaWNlACH5BAEAAAAALAEAAQA5 AhQBhwAAAAAAAAAAMwAAZgAAmQAAzAAA/wAzAAAzMwAzZgAzmQAzzAAz/wBmAABmMwBmZgBmmQBm zABm/wCZAACZMwCZZgCZmQCZzACZ/wDMAADMMwDMZgDMmQDMzADM/wD/AAD/MwD/ZgD/mQD/zAD/ /zMAADMAMzMAZjMAmTMAzDMA/zMzADMzMzMzZjMzmTMzzDMz/zNmADNmMzNmZjNmmTNmzDNm/zOZ ADOZMzOZZjOZmTOZzDOZ/zPMADPMMzPMZjPMmTPMzDPM/zP/ADP/MzP/ZjP/mTP/zDP//2YAAGYA M2YAZmYAmWYAzGYA/2YzAGYzM2YzZmYzmWYzzGYz/2ZmAGZmM2ZmZmZmmWZmzGZm/2aZAGaZM2aZ ZmaZmWaZzGaZ/2bMAGbMM2bMZmbMmWbMzGbM/2b/AGb/M2b/Zmb/mWb/zGb//5kAAJkAM5kAZpkA mZkAzJkA/5kzAJkzM5kzZpkzmZkzzJkz/5lmAJlmM5lmZplmmZlmzJlm/5mZAJmZM5mZZpmZmZmZ zJmZ/5nMAJnMM5nMZpnMmZnMzJnM/5n/AJn/M5n/Zpn/mZn/zJn//8wAAMwAM8wAZswAmcwAzMwA /8wzAMwzM8wzZswzmcwzzMwz/8xmAMxmM8xmZsxmmcxmzMxm/8yZAMyZM8yZZsyZmcyZzMyZ/8zM AMzMM8zMZszMmczMzMzM/8z/AMz/M8z/Zsz/mcz/zMz///8AAP8AM/8AZv8Amf8AzP8A//8zAP8z M/8zZv8zmf8zzP8z//9mAP9mM/9mZv9mmf9mzP9m//+ZAP+ZM/+ZZv+Zmf+ZzP+Z///MAP/MM//M Zv/Mmf/MzP/M////AP//M///Zv//mf//zP///wECAwECAwECAwECAwECAwECAwECAwECAwECAwEC AwECAwECAwECAwECAwECAwECAwECAwECAwECAwECAwECAwECAwECAwECAwECAwECAwECAwECAwEC AwECAwECAwECAwECAwECAwECAwECAwECAwECAwECAwj/AAEIHEiwoMGDCBMqXMiwocOHECNKnEix osWLDgNo3Iixo8ePIEOKHEmypMmTKFOqXMky48YABl/CbEmzps2bOHPq3Mmzp0dsQIMKHRpUI9Gj RY0iJeqzqdOnUKNKnUp14dKjSq8uzXq1qtevYMOKHTtWa1KzaLkeJcu2rdu3cONKNKsWrdYAXeXq 3cu3r9+dd+0KPrv2r+HDiBMrbrgV7+DH2BwPXUy5suXLbZFKhvx4M1DMoEOLHn1TM+fTnkmrXs26 NUWsWPFeCSAgQKvTRDe73s27d+vcRK+0qPKHhRXiVVgIwj1Usu/n0KMnbh5c44oV1gP8qWKFBXOh jqWL/x9Pni34oyxW/PkTgPifVOqxfy/6ubz9+/h9nifKItU0QQFMo0ogAFYh33xA4ZXfggw2mBJ9 6P0xjQECpJKFAQZodyCCCjro4YcgvpYgUgJICEMAqQjywgvtYXcbgpGFKOOMNBY0IlIsSDiNetP0 qKGLt72I20w1Fmkkgzfm9hJ7G623nkzzEXnklFRGF1l17Tnpnpbbseckk0JCJuVHHFVp5plkXSmU AOp96aabeXyQpZMBLMcZAGNiRGSeEfGJ5p+AgqQmUAJsiVyXxOXxxwd/9PDBowdUceiGg+EZUp4a CVQmnjPJ9JKmnWYa6KikPuRZi9qxt2WjH8T56KuRfv9JqWCWghRqrTB1imutmu5qqZ+lBissnkMV imgeyA2xxRAfDHHAARO8OuehnqHFq0d7cpqprp+Gmmuvvw4r7rin9tDDEHn08EcYPcTJ7AcNBNBA GFvAO8EWb9r2GLjYDvStv71K+e/AAY9rcLCnZvBoGM0q7OgEE8S7whA9nNFAtBQjyh5k/GIrqra+ cpSryLdee/DJZ3p2RQMOf+CoyxMcoFEDfwyRwQQB9MBwq18G4Mq+JqMs9NA0VXsFDhPkkfQWYTxh XQkfbNGDDLNtsSyj1HIcNNFcd01StdhEsVEktdSyURWjPBrAKZG9ymjPYVq7tdd0120R2LMF0EYt SZT/EIDSNn9wgJ3aZUBvly5qDexEixfMUON2R34f2DPoXcsfWzyRwcPPLidIFBkYjq+if8xqV8cX /ZsQ5JBL7np5YH+SRC1VPHrGBxF3wccpV7BSxRMfnHGopFbYCXTrGf0qapkDjywTrt6+Lr101UbD Sy1tQKz9FVfcJkgDN7Cr7pfcxU3r3BLtyW3BBDtOsK7Tx8+bZ9EMc/0MLFyRv3ettMIK1B/IALoQ lRy0GG8yyGuIp9wXLoFti4Hwk58EV+OZWmBjGBYUSv9kEAOr2WwLikJOFQRglv7piykJfJy/1tdA 6EGQgROM4WgGFRQMBuU2GlHYo4a3HfnELUhBYUFd/z6TQoWMiXnLc2G3mgeuIsrwiXqh4VFms550 8cxN2NmIkF4CFBYQ54QI/BpBnAjFMrpGikQR26vCEMIuqSoVdFpOqp40J7BhA32pG6MZ93gfNIJH h4/Cl5vmJKs5vok95osRHxc5riTx5wmjyMMoDMfDSmrsS5IcYoIYyUlhOXIoJojWqwQJp0M6KYuQ EkQiFdnJVgYKQvwR5aNsNopE/SEPyupSG0unylbITJVbcaUw/wTLobRCBgeg1xYcRrEhDGEFBzhX M4cQA40YTxDANM0wt1ml/RBFEIMIALuYxbCINWB3F0tnA2QgiCsMpkPcjGeNqIMULGhvAk+YwBX4 4P9ObLSTe/xc5VbqI8+ChoieSOHD/vQHoyQZ9KEfAk5DBeMciFp0QbCZKF2EctGO9jGjGjUNRz1K UvEMNKQSHWlJV+qbq9gRRkNkqUx5c5eXMuelM80pa9IyUU0SVKdABY1dfPpOswT1qJh5p01dStSg IPWplBHTRtLS1MlA9aq98ZRWscpVEJGxq2ClXljH6lWymrVBXz2rWkmT1rW6FTNtfatcFxPXudrV MHW9q173kte9+vUtff2rYMcS2MHiR6uITaxiF8vYxjr2sZCNrGQnS9nKWvaymM2sZh9r1sIOzbM0 Am1BRXsy0obItNxELbm4ptphtnZYr0VrZzkZWwb/1baTtyVVbg87W0budnJjfR4ff2sf4prxU3tE 7sGU29VNHfdjpYVuc42bMupSz7pQxG6VtBsd7sJ2s5k1EnjDO7/xXtaw6E2vetfL3va6973wja98 50vf+tr3vvjNr373y9/++ve/AA6wgL2G0gLnJSKo8dREGWLgBjNlwGlysITvOJehclFJVTWqVSbs YMUoVbEIMtKHEwsjBnO4wROhS4aTstQDJ+TEKDZvZU1C1RZf+E44kbFlSVJjFa8YKSaGMUpT3Jib trgwChFygY88YdQylcnNgbJKayJlCZP2yZz5sVWTrOSQEllJMK1ykLvc0Cob2MmxiZKYb2LmBl9Z /6RG1hqXydzQL3uzzJXaMJ1humfw0DilHIJyjvvcHB6DOcx5nrNd+teKRDDa0ZButKQj7ehBJOLS lq5zhQkT0hWPGS2O7l+oJ01qSkva0pjGM6FZOZIoo5TJg171JlsNnjZv9HR6tssgXLHrXvO6Fb0G 9q+DTWxeaxrBxVyy3BSNFl4729eDEHa0iz3tYf+Mz6tG8yc1euRYy/rN2+4prpldwkQgyNFe8EKI N71UQSinFSxoAfeyrGFya6V/506EINQdpW//2Y/crjeVZT1rkST71S3+tFkSce3vAHsQWFg3sm3a CkGsgDuBuPgKrNDPzmhF4VpxhbnnE20vDIJD/v8uSbibQ0KNCMA7Yvo4mwnOakEBvNbYmE1tBOpS mds7KGAgBBiGDoYvFL3oX0h6Fr6w9Bpkwek1qAENlmJpWmAj4t9JMcVno/EWSWoF2SzqUkAO9KAT /ehGT7rSmf4FqGeBBjVYCrCxQQt+3zTlPF6q/rYznO2MkN5jnznBr6z3vhsHOSw4ecwD/3OgEMIV tGC4IiLvCkWI3BWPh7YiBqEIF8QdKXP/WQuwfhrGaaVA1/G6eqoQgI6LHckvXsoXXAF5yV/e8rvO /OZ5vfksuIDqrfiZK7DQgrtn2yR6tw52rrP6KxxwqIwf+OC/ppW8pT4A2NlOdxYP5FwTZfa0D3// 5cNvecsnQhHo5zwhpL4UR/+MFjOYAXNMf5UrVGEa1JAU/qkBCO3UCfCwhxBX8XjiR37jh364h4Cc B3ftZ26QF3/GR2imZWN0Akd+xx6uR1HdZxO2tmTUdxXp8STuAR+l02Y45X1DQQgMZxdWBxSVx4Cg d3LChw3FV3p9YhYWNw0vQABVUA0YQgAX13qosYGxhxRfEHmC0YLYMAjshxSqNIOuIH9DeHwf6FL4 FwgBIgiBgIUYOIVbJn14ZytmUSH/ESCqoIWo4oVfKIBLAQa0xxmDQANTBxTlN35VR3fh1wJSCDSm AgBmYX/TAAM8qAo/qB7/p4ZOhYJC4YYNNxiK/yCHQXGAlZcIJ2d14TcDNehxVCgSMEEXApIhE2IA BeB/GaiBD8aBNAdanagVOTIhFXIhGVKCQxKAB3EVYJAISrgUlogN16YICvB5vDh5DFdpvIgNQUJ8 ONaHOLgCPbIdPTINpcN6QgiAU8aGRihyaLGLP8OEChAUkkd50TaDrUALo0eNeyZaM+GJ1MAiOrgi pMhzPbeGLdGBr8aJfqgVJRKIKKIiLGIgYKSJ8mgQVwF+hKB2bXdp2ACJbad2Trd+3UiH10YLtGBp 4nd+lqeHfKhAAlF9LyEpG9Em2/ESpRgYAWkjbUh7gzAIa4eQkLh0S8d0NUAIMBiMkZiSBWiRmP8I kH2Gjht5F9PgHwHiHySoESwARDpJYag4fbbSkyB4fzuiI9PwI7ZhlKZYjbW4FCroCl9ACFmwfmCw a9jge9hQA2BQA2xHA74IjEhodY0mCJSHi72Wk4mmkfdIFOEUgpY0KRenfQEAczxVkgQxkJFXkFzZ dsbmAllAC21XlgvZdg+JDZMHFFYnCJf2ls6Gka+3k/bIlJqxJEuyHtKoESMZj4mYlFFGj3d2F0vJ mbHRJJ9JR1OVmT9ljUcxe5vHeSqJdV/gAjTQCpuXdDWQknE4h5DZcLyWCFjgBViwnFgwA1cQf5m4 bCrEmkBhcZJiSnCyBfgEmpLil7dWmkV4FIz/+AWbpwhY8AVA8QU04AK8l3Q0kJIySZyVF4mUuJzK iQVX0Jzxt4d/2ZocpooD0ZkoQj5uQhzXcUrTCH3gWTSwgReNFRkOqmCx6VOFRSRIkTerUqBdYkga AY+pSXZAAQaWB5kpeRu4SQMKMJGuQANfgAB054vyqQjGaInD+GzPRnrSuRAWOhQWxyWWxDRVcAYK 0wBPMCnY1J+zeZVIkQXzSZ4R55u7lgC9yWtvNwO34Zje+Ia3UaMMB22uEJ3fiVCoOVSXEpi5oaGg 6SSKkgdhkAGG9HcKmqQsATYTWqdG4aAshqeOEZuasZp1uSYgaShcUkVysiqHGKZIGZ5EwaTb/zgI WcCLKdl5NEB7iRB1LWB1wxkUlkcLrnAb5jd+lzaMicCfAqejAVosknKd1+kkUsM0H7AzAaQdW9KX SJqotEkUIvozvwkU8Lmelwd3CUB3meqCk/eGoKoIojqMpEqSCJUgknGntband6oUe5obZXqq4DGo GkMc9eIqsZqmrPd8cGar8whS0qom0FqtEHol6vpSgSUlRGEseXlLi8IqkKKqsoioIIoNX/kzBfkF u5YFTEgDCaAIZCml6lZ5vxgUlombx+mCeDgIy+piRmSmQpEerOd3q1cIW1AIwXMDN/MqbpqmpbMC HkpD+8qkJ7eVWMBrXAkGUjoIZUmwYKmej/85ebvGcOS5a+gJsQwHpsx6Z+3KrvSxGXhKtHR6rdSp IaliSHGiLs0ism96qMzqbd4EJVqFrhLqrM9qrf2CrUVRgiJYRazirbDyBIUUp+QqkEvBpDIKsFiQ kjObBU1XAy0QrOkJATfLe+fHeX07FOcHtBS7OhYbRFVASviCL+dSL/ASAAcgtW/Ceki6r+S5jUmX kmSJmEsHrC1wbYOgABAQiemHrOWJkEKBrBN7Us1arXp6HhE6oSOStH76pwmyHciiS3+QAZ3QMDIj siUArgZyshVlmlJlRzbmtR0Br6iaB2iLLIkyBEHaLM8iS+3BQ0emG4oIdGBZkCrplS3QdjL/WQMI YKXBqAAxWp64WZ5QuHk4WqqEC7ZdJEq5GwY3EDWwGi8N8AGjALIB8AGV5J2qu7YmiRQ1YGxdiQWK YHRSZ5ab2wIu+jNYChRdynvpi54zyHmCG8CwRK1IexZHy8HrihWzS7uRcS7tMgSjYC6L4igZIDMZ wDQZcDH0ekqjiVBWK4FfC78lfC64xDTtErWNOy/1cjGkRCceir2NN5bohw1JN3RSN6lNt57pFhRx uLfmp4CbB7iKkLpEWLE6HACy5DaPgjMa8QSMGzOvMj6geb1WybZGuMRNnLmTCgZPJ4dTzKvmK7ro J3IKqHgtiLqyyWLOGruD8sEX1rUiTCYD/1wUaewyAVRODRAvtUO/EOMyYUAt4gpSN6yZybvICaJD DBM4DxMx2EcxFoMx6pI1YfppWQGwEEwDYMCbt0ELWiB143sbP+ObeUyHX5AI5NnLvzxy78drXEyL qKO8Q8EC+dso9Bs1TpNDPVAIbhoAT5AHuwu9iEJCq+x9WVHAJxfFv3cbNaAFCTADLorLTIyiQRHM 7JzAkbiEXxrI5qjBflYRR+TJOZe/r+rIMLMRN7AoEXMAZ+Aoalw6NUxP8zhGFsaniPwdXwVdYxIc LOPILxMGMTMzNXMzObMzpPMkwrugSsq1eOG2Ydl2kGh1UecCCRCsDfcFekvFwmlpMa143v+oCBnc xSuULToMFHljOXyjEb/LwibgT4/7KAUtudvMbFyUBSPKdLFMAy0YdQSbAI34uaHLqzNNicI5FIBc lRHIVMFkzx8T0UTBBxOQAZpzCfQijStQvxbNByawAq7KRhuayRk1j8tTVFn7uoP8ThSBXGQ9FEeT NElTBU3zNFEzNbNBL8yiMUymWa5MCzVgvtHmgpM9A5QJhYSQAFctFEqohO8XjBCXp1krwAtUuEIB IBpBNrVQAtjRQTYzAd5xBUVtOG1UBQdN2pZVA1n8iy4wd0DxiwhQccV4pQR7FKD9zphq0/LcZd6y Y4GdzBtRNmajEZijNthwCkXtMrKS27r/rWOMJXZcpKfTWrzktdNhMzZlczZpIyds07+vQrL/qLqs rBS8LaNw15sjl5A1kAD69s5foACPaRZKyHnFHK2Nhc/VudrYoxGG/SoTEAXV+but0tG0Gqf17RgA K6NP7GhBIXUJQJGRGOADTuB0KLHNrWRt9TzRHa1lkwQOHkA2cwDeI9dRgygr4N0QYs+MIzJ6Pcit m5p08dfdouA5pxF70zd/MwHvMji1+6r4YqDzTd/cLBSuDLEQq43FmMuDELOX56W9Fpnvt3kHXpLM Y+QAUja5m8KPMs3u1Ary8ii15EY/NryKuhmTStPGueV8/spUnbMUbKOVWIwYLM831mmF/ybWOo3e CBAAZHPJVSBAgnMADXAbLCADm1PE4WqKPJ4+e6LXWztVWfHYRL7o1AkUlbM3VqM5nOPkn7M5h4Og VbmvvE0I2Uh3xUjij2F1gTuXegQwqB0UylA2sNq4MAcgt8QuHf0kmZmyS6wVVhftkCrgkEELXa22 G6xs9fzXBdHiQdEGe+OxAy0453QKfCAIpZBPH/DPRnqyN9Lpd9NkZMo43W7kskM7AXQ7uXPuvfM7 wXMo2mfXmpy9QCGwIwqxUFjcxhiMCVDiuD6DWl55o30+03nq19PaMmE8fKDP3D08VYbEijoUaCmD 75zwubzwLo23uC6ZfK7l5HnTIHW1o/8OwqKOc1wLoUY7wkehDNeTPffET9U5L2EQ66AJdpoI7/ac ijl86taDPfekT93jT+AjPiRbPq9H64RACIqg9Vo/CFl/mympdiqZdApQsOZ3aZuH9jK98ku4xb7+ vtTJC8qAQYwmDfjWCitACmwaSDO8HSzQiKoL8rc6FAL79Vy/eYTg9TEt9mSvABaJrGof+VhO5oFs tISMrkCO+Zp/tEWh80Mx92Ujb1GgP3xgjKxgCkU9BOx+SgIfFM8XHtyeOkrfycFeP/cjb/tjjP1T AgojQMNTQDjoTbTueXD3i8aPoshvvlLa8Myfogyr3H0ejF9Q5m18EMgcFNEQDcMu2Ff/4AChQzGH Qy1RrxVYIETFtK9wR/yTvf7m2/7Iv/xlX/Zsv+cRz8TM7dWOxLr7ka55ivnVUqEAAUCgQGwFDdbC xguhQWytWgXYsiXMhyF5/lz8U2WFIIYdDT7kWDCAwYElTZ4cGADlSpYjPb6EGVPmTJUsbZ6sOdAj wmELP7ZyECPikA9bLF6sUoXFzIIPWxl0ie0mAJk0aljFSkOrAhpcFUBQEFZsAgUJmBqkhS0tIUUt zpKcCiAn1ZfRlH28YmVFlR4fehjFeHFFABaCrjC8kvgwNix7o0aVelPmVcpatULomvmr2LEK3qYF rYjQjLciPT52OTI1ttSqWUNluBqy/8i4teWahMmLF0NBAQIQ/ZDhj8UqGf8ECMnw6dOCV3wzf13Q Ns7pK2fH9O06dmmmc6ub9A6z50fWATJ8QF9cfcbBDTsuN8jCN2y4NrljG5SoIC1Xifr/98+/QVwZ cBBFCjyQwAMVcQWbBl1R5AssuKttrrN6iyGMG8LooZD1ittrBRGrmLCV3kS0QpBBArjoMYamuq+V QQwKEMD/EMRxwQG/ONBBH3l0q7TrsmsNqtayew2y+aJTbTbvvgMPt+5YHM4v4zCqgsilmsuOI+eO Sy4qKAV68rvrYHKtySZFIrLIs8qEMjym5Lsoj76OunKwLo1ckoXiAmAOMhi5S2TGgv/8c+U/BhNR ZNFEGM0P0kIPLNTHtHgkrbQKS2KqlS4OOC+4o4oTbIXAjjORSsFUXe1FybjLzyBFZ2XUUUklNTBW B2mhJcJM3zrzPmHRHLMlTpkyIQD0PiDuyixTwXKkFkxdlVpA6YsszmJTAna+JdkEF8lvu9s2Sp2Y ikJZ9MJoltQASG2RI1WPe7dFbLNl6b5BWklrP7RkBbigBtMa2McHHfSVwk0JmomPA4bIgKghPvxD RKSu9M04eNXbcrZBS6u0X7X+FThgSw1GuUEguQt2WJehKpe6c2VSd9lTK54XqXpvjhfbYuGcruWO UIutSCeBjTmlY2kOFb0tbt45sMH/eMbo2uikexVkQ7GpkVGv+3t0QVwLJXsQXn0Eoy2F48qJqSuo zWMiqucODM9vPc76LV0d7FoRrymF1G9K8yM8EV4JJgSMX998ufGXgB7TQpqGq7JdjS+nuE6cw3RV 23KF3s40ccvztnQhkyZz6ZhYeGKUPEbJIAyKZ7+87lHaM63zfGE13EEbBSRQQRwV5JFHAw+GUMK1 2WZYJkEOWJYiy4ubHmPf0FPSo4/fknHkrm8MPlfhiY+wx4NXFtJx9WFGXenmsQuVqFExwhPeaM0b ogEohta9OsgrXF8AWdO+27wPJiaYAEXO8zT7YYlqekLPAQQBHbzZh1BbY4rIGNIg/38R7GxgGM3y plITtzUgeusaTh4sMhGLqBBPxzES/zqyvbPs7SwaNNnhdtUrxbFMgI77n+foQrMTxq5ZeSiEcI6I lKW0QgYB6AJ0roY1If7sh+oLYtBmtroELgtixmnhEBhIOVKt4AoOOQByOJi7+uyuNPsaWPCAVyDh 6WhBxisf8hKmKduQkCm9oV4hTviBBlzvhEiimfbydpbuNQh4AkpQruyIR0qiLG1BAtYVX5ZF/20R Jq04QAPCEBH09KAHQxjCDQLQAFSecgg9CMAKrMAcE0lxivjq5Oc02ThO9tGAL3HiAUa5hb6Y0pUl OIArXRkD3yRHEMlhYxtXcp9KHf+qP34DW+DGVjjCDWJGD6IFCDF5lk5yjwUTAE4YwpAB31zBFQ9J YxoDIA2HJFKRFnyjfgT2twCFzUDbLJs3B7QrV4Chh+nbpbB6qcUhOu8KMejBRF5JyAk4wDATaEBG MWoYW8qwfwzdFugSih0CFvBCgygBh4gyEYw2gA+nyGgDWioDw9Dknm58iyB6VzCejgx5KQPqgxQB hgnxMWil8VQoDzAYGZying2ZYFS7AxMadoojBPNpVrH6U66iDH2ZHCnLSpq6hs7kChNAawNiMAE+ nLE5bU0MNLFz01yGNKwKHatKSoMFtKL1CRO4QlsLYpjEtLWjp3lJVWdSzZH5y7H/jfWRwM7misNR Ni0GXRxTzMQdEz3Ts0Ck6iKZkh+RlZZkkUWtZXfIH14NNbM2vSvSxnob7rDgCra1LWjpClIrxvZ0 JdVraW473C3xMrGinckEe7rVoCLMQMXLI0H3+JY4+fZxMVGsTBLBkeU293x3LN8XDBZOtSHUujSb LZnOi9jd+lKX66XJbEU60mBlNybeHBlo8rtf/YKGPwRlLWUpa9BxavY7t/St0OwLE2/2l78P1q+A Azzhgh4UrPB9XHrVi+H6Ri5m8z3vQo/KYewi13n+8W6Ku4q8xBWVugdGcFhBt2BgclfFN0bef79A iALDFsOn0bBcSBzaKkbuxySV/++QiYzTszBWZGuM8GQrO2VeJWrAFiYnjGO8S5HS2CP5gXIOpezB Kk9ZwInCrA+PPLQgbzmhCvbwe9fM5tm6mcsz8XJHwIxjHP9XwGb4wmtlAqXQyTjL+HxLpVbM5xyD Ewxm6LE950wbDUfT0IOOs10nHZv0Wpq+BkZ0DWfkYP9GGMASRnWAKxzpElfHo3c+NJNHazYIq2Wy 4021lSmcZvNOWsSu9rQmu5zp3m6a0mON4aXxbOKY6LRBrQBfjigZXmqnbcfWNqiLY20b9grbqLJe LHdpMcdIgve51IYutrHNajQZe4CVTjasl13k6rr73cgutLe3jRJqmo0/iWhFgP8ataBGGahRPGLL jtliUIY3HMugns7jQNwtEfKbd5SFNsAhWfCBGyjhQ40QIRbucIYLeq7G/nXEXx3AcWGa3pu1d8rj 0m0Bttzl4D7xDLwwAyy0YAY+B3rPfz70Fggd6D83OtGFXvGZo2ni2Hk6FXEek+3uvOdJD/rRs050 pGu962retMxrI/GaM31hmna32JFr803eJ8/2brXKoe4ytu/7JHB/mVzatHe+993vf29Tm50OxKj/ LGmAR3ziFZ+dTg+el1HHpcXxLqz0EmmEjIf35N0eZM63zzdm+nzn3Sd60pcePL/GPL41b3YCIp70 q9+86WUfNMtbJ/WmV/vsdU//+9ufPvSZh73dd8/54H97+MdvSd+Rn3vkN98mfi998V/sfOJLH+LU x372R6997nef2daXpveTBv55i9/8s2f++dUPfPK3d/1abr/73z9/z9Pf/siP/5LvL/f8f3T//385 ABRA1eu/8POf3uO+ApyhAWRA3mrABxw/BTRAB9SW9Au1AoTAmwg8Chw+C8xADqQ+CZxA90qJmoA+ 6NM7Mvm9bRFBqftAlEi5J/HAunpB0ROxGfw+8vuc0DPBAurBAroNlfhBIWyfFoy8GnQfvtO7nLA8 JUxBxkPAw0NC0qu9JVRBFeRBE1zBCBTBzymJHhzC1IFCHxxD1DHCKYRBslIa/zDcPiIkQzWULzTs PB4EQh8MQrJyw5I6w5D6QjHEwyu8wx+Ew3LZQzlMwhJ0Qy0kw0QExBSMQ0PUMDp0xCZcxDwkoGEh EoXaJS/8wjCsRCHEvDLkQiHJREm7FwGCxEO0Q0F8Qj98wz7Mq1SsPFa0REbcviLUxGAzRVSExFxk nJULIFm0wyVkxDFsQkXUQhy8RWE8PDD8PWNMxjo8PiakREu0QkfERmYMMia8RkpExEbcQm2URWUU xweswkCsQza8Q2ksx9b7PFt0xVV0RXJsR/Gjx3oEwCF8x3X8Q3XEx0dcx2jkxoBkx39Ew3s0SPuj Rj+sxXkcxoSUQof0R3WER/+IrEGEtMj1o8ZQTMduXMiM/LA1FMMyBEVFbESQ/ECMRMmVXD6WdMn9 U8mXlEkbjMmZtEk+vMmcpL8o1Mme1L2a9MmgBL1wFEok3KRS5EUaLMqlnMZzZMoXZBmkZBOpHJYA fMqrpEKsnMI3qTungzwXPDutFEsbHEuopJmvjDfjC8uyZMtYFMrFg8u4dEo9dDwsejqrbMu8NLy3 TDvBozmWU0vm0cvBjMigREth88u0/CEQw0vCdMwRKsrDvCKEBMbJHDalfMzMFEzD7EvgExpBYIF9 Ga7F8EX9W0vNRE1j4UuU88uWaYUrSAoreBuNqAK56jX5e77U1M3cXM2wEzz/13QOERkMjWCPCcLE uCPB3VTOZcxJyVzMNhMa50iK4dyLpMgS0gQ73FTN5eROoLzIzuy06KyCaaCGvSBPVQAEFkEOujPN zeRO5fTOF3TOmtvGmWiMaXgBAqgCVTAAAyAAUwkA7LyPM2nM9yTM+ExJ8MwrtxlPGBiAP+BPAygA 9bTN2wRL9zRQlyzEDJxPlovEPxrPaUiKaSDR48iSAG27BcTMDEXJDYXADl0fegyumHCO7HAM39iL jMgSAWin0jxCyGTRl3RRc1TQ1isrg+gNP9FR9bBOeFmBElhSASiuXxxB3gxSlhzSBoRRLHrElxCE EHEg2ombBngCpNAIdvvL/wK90nrMUgbcUiDq0o740gY6FYuQCDshigagFsHAggq9rirdzjUFyTYd wJZBpNgiRz+KjSZdUuOQiCqYiIlYFhZZjwBA03xTU0EVR30Jn/CREQL51H0pkFAN1aTk0HbTxcl0 yyP1kyxRD2qJiC3wixvIgKZhp1NZARY4LJrLVE1lxhgBuIwLuGHVOGId1n0BuEjZ1aqUzz+NN28p miMhndwZklU9jSrYAg/5g6d5mlO6AfSYgACAHoi51fXICBATEzPx1UGFlTVqHBkZBG3Lu2aluVZJ E9jIHtmIMWVsm444J/T4gwz4i6LoC/QoJBOKm3BlllPJkgvr1XVNRZkAof+JBYMvqNiL/YIvyAKN /YIa6NgsqIGruK/ekVdmTdB6FZ0k2Q41WRJ9BTLg8iTWaJpBYiffeAJZ/YAJgB70EI5KdVh1hViI lAlC2LGiJYQsAIMsONqkBQOPddovoAGorYGYIBAH8QL1+c5ua1mVJRrt2NryIBaY/SVsiAIT+oMh 0JAPCIMnyA6BlVUc7YRO8Iv6EYCfXdGgLUeZ+AIr41sGSRQGWZSBYwtCEFmYADjQ+LnGydrTaFl9 dZEj+dotw0HJ8YgaDYAkqIVaSILs+IO+mICleJ5l6VkdRVdAtQ68Tcih9Q9FGDcICR4GGRBIkiSr iIlGwoaiU1x61a2pClv/nIRDmOgN32iDzC0B34iBzs3ZpXCYZemBUVEKuwVB1P3VmACDRBkWwqWB kfUvVzA54SvUGOVKJAspSRxbbKjRSKiFNsBRSU3ewQoAQSKj4xDQs5TLfZRefBxa1s2ggFEEF5ha gWHd1bXdp3CFnjPZU2VNLKzfDSzf3sjcP/iATgCOml2MVUKPPLAfwjCv+/VJvbVeYcFegwBc192z KuPe7r25F/XKIXFWuvswBj7SglCGzD2D6CkkEyiIU2CRPDiD0XWXATVdDmZXmCAE/9gxpa0BMBio LHABbEhij91YGlAEBfhftdCPtHAIswkQWohdA2Y9AQwWF8m3U8SrHWTF/5foiVooXiJJjiswoVKi HJ3xRSHWSQ8WjVxRHmyAWhpohQPJ2BoQKK0Q4TUikETAAi/AgkTGAp6bgcQNTDdFVdJRk2mlZETa u97dy9HzCF6Ihp4whYagp6YoAVWAVIoYIxDRVSq9GjrOSYk14gTBAvHChiygARfgYkXIgkGgAaKd 4iqGkJ8QBETeuWHmOdx95O9FWSZZ2Wiy12Dj15iNBmyY4WhG0ihwAIF9pYiI4+OYJbeRj0Jj5ZuU iSywXm8SLwIxEBeggURJhJANEl32jIJgEP54ikVx3UfpJxTWzv87EyUZHUlWZqS85Jc10vK1C2p+ Tb0gpuDQ1vXQk5oyiP/EwILFmNPsudBwblHqHSii/QL/yOWrSABFSOIEoIGrhRAqRosaER9diaNB 0GcVRWCP2tpmdlmvFZpnLl85DYANUae/eKGKEU4RCYkvFc5nshZwxmiZtGM9JgQs6I+MZWIaGASQ reWQ0GUIGOQRjpQeYQhGuVSYVmHGLR3TyURLNmtwyTCxjeH3CF2eHQWKsZhoGYTewJKp+ZN7SWql pl5CaBCiHYQdS+IW6FhCoIEsQIAZIOApzl4A1pE6IuQDKdkUJtIEFlu3eRhUkp4rAeqLqRpTodOM qNtoymshjQlyVgQ91ti/Dlmr2FiQNguV6YpBliSOwyCuUYSX9j9Ipuz/gnabFWihdcKc4KadSrXo ix7t1NXovkZiMCDpVmhaMFBnLBgoVyAErkjpG4kkfWpp3A7i+3tTXopTmeCDN0aPiuCZnz6V7MAe aRjj48bSmKiBRpnljq3lkQlZF0iA1zaILwALg+gmgfImxjqU8pq+mPa1yltrhijbQboBylmhhW2h i0iFhsUGabC5CnLvjNTbgdJY5ubjhqBqBGgBAkaYBMDqQymeREhxHtGjeP3ifCzS+ruQeskDnLVh cR2kQ43k7s7w6YWJ+D5trAg4g7CKBNiuf/mCsCgNkTEQ7jZuMI5xz+MeKECnUuph34gieBLX53CP XfzRHmfTmPgCAymZ/5JhroOREZKOXTkaHwPBql5x6Rfn5yiXckaSgQPAczxvKqfKYaeSKt4FazD/ x3EWjfu4lLC4Jn4CHK6mEQL3Xijf7QXNKUHgA0HoUz/NTh4X9F4U8wEh5Kz6KQImBLLwiCcDdeeK bOScbN+stO+20CffdB9/ifhmC487EJGLkL/+64zd9SQPaVtZdMKBrCaXc5ikczO0szdT9VjP2063 o3MrHl4e86IdBLJY85W2o62arkeH8UinS1f3seNidvyNCayoDMvgCszYDM4oi7JY8pFhi6/ObWRm 9cZL9sUsv3Fv9h8vbHOn4n/vihowcZImi4FXAP0qmYM58whJ9fbU7f96rzPFlLdl13dh7LdBZudG uSZg32puUjR4v+1i9+5jH8VUBd/rq3hZryF+OS1TL/OT6Sk9CjSRV0iSj5kWxnduT3k0vPgAIrVw 4jGa30mbJ0Syq7mJ2/l91xvt/p1OxXZJOrfjMRiGF/r5A3dMTMy/pLvCS3pt7PkbOq1PJzOmdvIv h3SIJ0CJQ8uulOyuN8T7UC6TEXuYx7FtR3lzXOC8R8Gm5CSedPshTrStEThF7/jBKZuy8TPykndN D2fF+/tW5g5no3uFZ7SUGaqGF/fH1/ye/HqGwKHHerKzORzW4hVeK/DNR/2Z1Bdo06q6h/boEjC7 z/fUp/2V1JdRq7X/nze1XOOPVav62gd+i498FFu04lexFvv94Fd+Ofz6TzfzHVot1bKs3r+ysl/+ 6zdIuGd94+f+lOEP2W977Bf/iIUV3C81/vKzU6OwRGEtAkv+8Yd/AYR7FMs4p08Q6ML/aouQit1/ onr/+AcIAAIHEixo8CDChAoXMmzo8CHEiBInUqxoUSK2jBo3csyYaBAtWq4StUo0UhHKQSgVqURJ SNHLL4Regqlp8+aMjjo1Xuzp8yfQoEKHEi1q9CjSpEqXItzpdFCiGV5mYGkxwyrWq1qzct3qNatT nUzHki1r9izatGrXsmUY9i3cuHLjtq1r9y7evHr38lU49y/gwGL7/xIubPgw4sSFBTNuTFcx5MiS J1Ou7NYx5sw8LXPu7Pkz6LKaRzsObfo06tSqD5JuHXg17NiyZy92bfsx7dy6d/M+evt32N7ChxMv vhA48o7GlzNvTjs59IzOp1Ovzjk6dOvat3Pniz159/Dix4v+Dpw8+vTqf5o/v/49/PgJ2/+Wb//+ e/q38fPv7x9AAAH+NyCBBSIVwEAIGrgggw1CpGCCEAoooYAAVhigghk6uCGH3UEIoEAaIigiQRqC OGKHKaro3IckohjiiRiCOOOLK9p4424UwohihjzuCKOFOAo5pGwyBnkijTTKaOKMRDr5ZI4PfQgl lVWmNmVCFVq5JXiXXXr5JZhhijkmWxhq6RCWWBqUJpltuhmRmgyxudCcb9p555oh6kihiBP6WdCE RxqJJ6F2vujij02SqGiiTBb66Jg9Blqjj2YqqeWiNUK6KZiO+hjkp4fmiSSpnJrapY5IZmrhn4Lq 2aOep8p6Z5yz2norQrWSFxAAOx== ------=_NextPart_01C52880.DC295B70 Content-Location: file:///C:/C9344E54/Paper14_files/image007.gif Content-Transfer-Encoding: base64 Content-Type: image/gif R0lGODlhnwIMAXcAMSH+GlNvZnR3YXJlOiBNaWNyb3NvZnQgT2ZmaWNlACH5BAEAAAAALAEAAQCa AgABhwAAAAAAAAAAMwAAZgAAmQAAzAAA/wAzAAAzMwAzZgAzmQAzzAAz/wBmAABmMwBmZgBmmQBm zABm/wCZAACZMwCZZgCZmQCZzACZ/wDMAADMMwDMZgDMmQDMzADM/wD/AAD/MwD/ZgD/mQD/zAD/ /zMAADMAMzMAZjMAmTMAzDMA/zMzADMzMzMzZjMzmTMzzDMz/zNmADNmMzNmZjNmmTNmzDNm/zOZ ADOZMzOZZjOZmTOZzDOZ/zPMADPMMzPMZjPMmTPMzDPM/zP/ADP/MzP/ZjP/mTP/zDP//2YAAGYA M2YAZmYAmWYAzGYA/2YzAGYzM2YzZmYzmWYzzGYz/2ZmAGZmM2ZmZmZmmWZmzGZm/2aZAGaZM2aZ ZmaZmWaZzGaZ/2bMAGbMM2bMZmbMmWbMzGbM/2b/AGb/M2b/Zmb/mWb/zGb//5kAAJkAM5kAZpkA mZkAzJkA/5kzAJkzM5kzZpkzmZkzzJkz/5lmAJlmM5lmZplmmZlmzJlm/5mZAJmZM5mZZpmZmZmZ zJmZ/5nMAJnMM5nMZpnMmZnMzJnM/5n/AJn/M5n/Zpn/mZn/zJn//8wAAMwAM8wAZswAmcwAzMwA /8wzAMwzM8wzZswzmcwzzMwz/8xmAMxmM8xmZsxmmcxmzMxm/8yZAMyZM8yZZsyZmcyZzMyZ/8zM AMzMM8zMZszMmczMzMzM/8z/AMz/M8z/Zsz/mcz/zMz///8AAP8AM/8AZv8Amf8AzP8A//8zAP8z M/8zZv8zmf8zzP8z//9mAP9mM/9mZv9mmf9mzP9m//+ZAP+ZM/+ZZv+Zmf+ZzP+Z///MAP/MM//M Zv/Mmf/MzP/M////AP//M///Zv//mf//zP///wECAwECAwECAwECAwECAwECAwECAwECAwECAwEC AwECAwECAwECAwECAwECAwECAwECAwECAwECAwECAwECAwECAwECAwECAwECAwECAwECAwECAwEC AwECAwECAwECAwECAwECAwECAwECAwECAwECAwECAwj/AAEIHEiwoMGDCBMqXMiwocOHECNKnEix okWCATJqzHixo8ePIEOKHEmypMmTKFOqXMmypcuXGDUi3Aizps2bOHPq3Mmzp8+fPLEJHUq0aMai SI0eTUoUqNOnUKNKnUq1qtWLTJVm3Sp0KdOrYMOKHUu2rFmyXL1yTZv1rNu3cOPKnUvX4NYAa/MO VTu0rt+/gAMLHjwyK169iLEdbkq4sePHkCO7Zbo4sd7K2CRr3sy5s2eWSTFbzlv5s+nTqFOrHohU 9BUWrQQJaMFidOu+q3Pr3s07rtHWVapYCR68he3fmXsrX868eU+iojWukP6nimjLh51r3869e0jo radN/1vxR3yqANZbHd+b3Lv79/DjA2AfHsuAVFkMwAjwJ4D69V3JJ+CABOrWFWXTvHDfCwYYsEJ6 AAZY4IQUVvjYgaGJV0V54vGX0X8ABmDhiCSW+JZilGlUxXQZ/eHiRrUdJ6KJNNZoI1QoEnVFRqm4 6OKGP7pYApD9GTfajB7JZBGSNzbp5IQ5CnUFeT5WSWQeeQxxAJFVxJgYAExWhOSMY2I00EZociSQ mk+26WZzUbIA5IZ0VmfnB398oOeWdkKIGJhJFiQimWQCauiaYBZ66JuMNrpaZVNOVx1/PvbwwRZ6 Zsrnj9dttaiYaw6q5qijhioqop86quqqmkE65x95wP86xA1D6BnAAZrWWcV5X4ZJUZlMlnqmoacS y+qxyD4G6QR65plnGJZ+kMGtH+TRwwEZhBFkfyCuhepFSiYa6pmF0kTTt8mmq+5cy2bq7gcHZESe tNjWOkSsG3b6Varr9uuvfJjxMUEGeWTQyaUlZITrFhmYgN4fQ2RQZQCu/Mnvvxhn7JxrGtVSSxIZ xSqtCdhEkYGe91bHq8W+auzyy7pxHIDHGT1haa0HSIONnB+EsQWn3Xp7MUQtCzr0oDAnrbRPHEdS C8ET1NpAADEGcHIYuvr35dAOIZ2oqOWKmyZHwi5t9tkpiTaM0+5OLQg2rZQAa89WeskyuMaKC2ix wOr/XSzagAf+EWbK0LzRf67gqifWfdr9Z9FEC1vq31/vXS7kgmeuuUKVDaPMMLUgdcUTf2CKZ5Bd rnUd1w317bfl4+79Opub1277QZVFgw0voQ/VQgDMfpAyneddwVUrfAnFOkPAkj252DJFT+jt1Fc/ n76KVYGlpbFOrJGXGhk/iIdJLR9Rmdanr35EUYYWfBj4qvzwj7VR+geLL5aP+UTor+///wdpn1Hc 9bOJEYlTVUrgdfYHwAY6sCMYas0o8tCJHjCuTvHTVZ+qYx3KPPCDIPxIBAeYqZ91L4Eo7E8AGvCB CRgvNCGMoQwlQh8SRqxZG/pZtjTYQWwIIgAvhOEM/4dIRITU0Cik+0O0NGWCk2VqWtgDTxGnSEXw JCUKA5PWHxp2mB05zARRSAxeqEjGISJHdCZIowDCGKEclfGNITxjG9miPDjasYG3meNdGHPHPqov NHqkTFH8SEjrCTKQVuRjIRepOcNE8UjlY6QkBUfHNiavjpPMpNlU98hKRlKToITZZTqplCiG8pQu E6NG0nLJT6LylZob27lgScsPMrCWuFzfLXPJS+rtspfAjGUwh6lLYhqzer88pjJdlsxlOtNfzXym NJMVzWkCU5bYzKY2t8nNbnrzm+AMpzjHSc5ymvOc6EynOtfJznZmc5HVfFI8vTNP7tTTgfe0UT61 s//P5vTTf/8sUUCXM9DeFNR6B61QQnez0Nw01HYPJVBEVTNR1FQ0cxeNT0ZNs1HPdNRsaIImjUI6 IpLacZbpQqlCw2Uhk8KRpSmlXUllutKPgtSm9sSpZGhaU0LqdDs/3Sk/3elNpRG1qEM96jZ3qtSl 8tOaUI2qVKdK1apa9apYzapWt8rVrnr1q2ANq1jHStaymvWsi0SkWveF1q52cydrjasi25pTbJ5E rngFCUwTsleX4FWucDnSKueoHcG2cjQU5WlMDvrXuCYpnopVSWPXGlhODrYraDpOYS2rljRp9jSR nUloaThZRIIrn6MtSWkR2VSnguQuh20NKeeattb/avM7jpxtbNnamdRy7rWr1SNE8qjHRxYmuHOc 7VoHmiIZKdclyl0ubmXr3K1FzrayPIlvmbdP5BL2IXIsLleOq5dWJKIV6D2ves3L3vWedxCJOK8l vQsdkQAyQrvFDUuiq9aA3jdEpDwffetbkoHmc8AAGu4RWesp+yJmEBEaBIRDhGDFOHgv/DVMXqBb YQsDF8OB7KSAK5zQg94zMelFb3vNG98Vw/e98W2jggXY37ZcuChgAAMhvpBjMPCYx1n4gpCzQOQv 1ODINchKIrDhBSysJ8PFvTGUYTvelkw5udOlsSVNSbQOm9gk2w0gYioGoFZ4IUIKvvJleCvCpHxh /8dwfnMWwDDnGtAZDDQw8hfyTIOsQBgLZ5YRiS+sL0EYRxBXuMLbxNjg/Xp5uoW+gisQfQUnM9rG D1Ezfkfy0HkypceEyPGPfezjLwS5BkTOAqr7zJQ/B9o2w9X0mpFC3qKE2hW4zrUrFLHrXSvi14T4 tSIIkWSmJIIW2KDFDASNYOZqWUoPsoIVHrSCRSNGXxweNHD11YJpT2lFVQjirF3ZukeHZKLxZEqc FfFmOd/ZzjXQM5+LnRRcN3k9XQ6uaGpNlDcvWS8VG4QrXOBnbFTMFVgw0qUHzFx97SgALFoBeVZw haB5krYokfWTv6MvFshLXhMXN2nYzDxzD26/E//5NCF0rWteu1zYgxi2IlidlCW7QtmfLfdariCA 2LCABSLHziBvTJQckxkxw6Y3Uv6N6xksW+jNFgn2AtCjB/UoFeTR2sIxrl2TJ+kuf8C6i3oUHK2P e+iZ9jrerExapOwY2WNWxCCUXpREHBwbLbC0dUueluEMh4MKhyTXLcKUNysiL0f/wsyHogha2L3x b6tYK2iBcKgz/NxpmYZsUhGILGDBQxa/OCZXonEZ6TXzgRBAKgSBhc9zy/KDn4nalwTdtuMYDEfX S7BpjhQII9sVTkds6+YDW4hTx0+CH32bkeJjV7R7EEL+N6uDfGpCDILmR6e83XNt90Q8/ezeDWj/ FAOQIALgp0HkS75+ST97MW0FedPQwoIMUIAHmf3atM63VgQrxpO/pJlMwW65txUHx2705grH9nit cHeTF3gbNnwESH7jkQrmYX+hN3Lrt3xF4W9wlgXERmYERwumBgaopngKMBT/Bjc+5HgI2IIOKHr6 plfEZxjTQA0skAqqsAqBgB5UYxuYkW1RFyiZZ4OpQA066CGwAXvKx3fIgRcboRhH4VlRCIVP2Br+ 938zdnuuIGHQZ2pkNn1D5oGLJxSHJxQ3J3ALiGt2930PuBAzknn2cX4FAHpPloEeUXhf4HyDEGzQ hw2mRgO4Bn1H9mbENhRHp4aA1nqV5nTBh4GX/3dyW/E74EZtwTEpQMcC1qY6aMd+2tYRIrIVHvcg wUGJZfcaiYZ/sYc79zWFTrgUmcWKUKhl+4RTyaRuXyCCHfgF/2Zp8TZnNCB39AZ3yKZiLGh3g8CG QuOGAvF+5Cd/qcAg9Yd8ycdvQzFnZagIWPAFrSB3ipAArrCANDADCOCHinCCZAg3cCdwxriFWzgD L3hIj+iJy4gUggB0xHGPlfgiTzCKeXd2QIhZ+QWDViiPM0gUhiYno8hB+fgHJSCKVdACmahhqWgQ 44cirXiRqwSLUziQawcTtehmuLdj3OhkQuYC32iANSBhhEBzZWhwW9h6TXYFM4AFjKgXw1eQGf8y geIxDR5yfwtHjUIhZGQ2ZM6nBTVAAzF3lF7wNnKHfS+XCHInYUShCC2AjBJZSo3VT29YFK1AbRuk QdZyAD7yIEH3X+3BifTxhBdJhWuZkZklRGtHj9NWiTxUHaMwAQ/CQdXmiGepcxoWkCmyQARpE7+k blkwlF9Akk25gNCHADNAC+yGfZJnht3HjgJXlTapjPMYmOjBIlSXL98jeECJDXP2b9j4BXtoZ944 CGDgAoOgmOaIDZUpcAKni2aIDa/5jmaJWQfillEyNryZI5gxi6xhkFTiI1dSOj1gLfQyJ6njSf9I hbx5GKxInU5okRbZKahVnEMhCH+HQhsSKz3/oETNiZwtcIFHNGJBSHs3UZhupnhCQQjZiJIuAIxI eWYxF5vChoDciJpT6Y6Z+VubqSMtkkIGxCmOQ0ej6YVBmZhzRwMuEGR75o0GJ2HYB1+/BpWKkAhf MBSQ2YiaKEfWeZ3UiSElSqIcyZ442QIrkpf8EZ7Qkie1sid1Mh2OGJ2LkZHS2Yo7ip1VmKK/wp1C gQX2Vx15GSu0UgXSkikBkAd00kOVpJ7xCCqCEmboYjR8BV5JQWdD+ZquUGdIqQhKKXAG6KHquIX9 aYiugJltmKUDKhQsijpfCSujAC+6opv3NZrxRghCAX3ERgPqAW804GQVY4KMx47GyG68lmy4/wmg fDlC1omdJqqjJwqkQfqmU+KksAIkPbAFlnItLGQrB/Sce7SEGeeDyXNY0TGYRTE6mmon1uKpHwCq uYKcpHqVUkpf22kmiKIkwWIm5xI9bsoQTBFkZTgIQYasLkAD8WZk3liogwABjMeNUBlzUomCjpqM vPotrXqcCRQreYApeUKjyOmTh/RYQtpvWVCG7daauEaCeGZt0CetlCls1XqtZMimIZpIroiisbiR 1xmLlioRTNKqN5AB95KwWzCj8BIAnaolE6ArK1OqfZk27Uc0BFEUfNAAQ9ADoyAxEtMDM5oRE5An DVCy8fN6FJurjgSPR9KRw9Kr4zIm/SM7Yv8zEw7xaTVAZh7IYwlAC60Ar65ZgHtmiGdKmxxqcMm2 plZJcpVzKEVBpNYxKUFyMkrqLk36pKSUHZ4oE0wRbxNmZK2pgkdZlepRqN04FLqoeGvLbnantLkJ fgA5nZKKkZUqsEYxmDh5BcFjL5kyAQcQL5gyBCerJ+PpI+aaR9EZfjDbqn27pHrSAIE7JMJTuLN6 Qo+0GJEza76Zo5eVFgS7rTFrs7PEJoTiNSo1ugsBHYchZOzKY0e5gLBLA2cWcIOgn7/Gn7/mn4zn qOiUrlPiIieUJWFQK9OiOHaqQBj4NelUrOsqFDQAoSBCAwkAkYYoZLGpF4oAogIpt7u5F/z/I7pc ybeXUr49EAMaMZ4fkBEbgrDx00lcq11jcrECii5IwQInUwg98wFnEAPxYjVbJLgLKzFAkrilNDZ8 tbVzK6n/ehkEWzaji7phQjakSyyY804Zi2EZYWQ8S4KACjdZsKzjKHk7xpJte8KMimv6WmVV6jVv urHNgic940TH+wEfiy2FUCcqW6qPtZVIkWcQdogprLR3h5qr2YLWapmLimzHiKfhpYShoZ38A8FI IQ1XcAArcAC7cAlPUDOYci3Y0ESWIjGIO4B5mnHRQxrDOUIvez5ULDoHUAIHUAu70MXooaQZwALS wAJRIzyYu7wUmU3YUZ0gtrXnhDukMj0z/+s8iaw3FJnA2oRqZVgDy5qC1JsAKeiHX5C9eXGMTUtu 31KwRCEw76Ip8jKuVmMvfxyigZKu1VgDLckVyEYL0KcA3behCBhf1DqVK0yxuvq7rowND5cEHtOQ AbAFmOJCJcMsGXAGdQK/2OVZIeqWUfiWamxOGdyqGVECtRAJIHMAMaC/GRBGJnAz40knSaigmVZ6 PvjAFKyZ2pWzIOaHv3ibt/l7Snu27AYBs4mmtCl3SguZvYxpw4oUfEABBJMBH7sFCQNxH1AIE+Aw 1RExdrLDEjmaQGyIGn13k6l4CeChGn3PcJtw3rta1eSr2Uyg3NwJtzLAGaDMrSADnmolCv9csczj xmTiiITsmw6M0/YrOgHAzXPQ0sY7znAjA1XQqX1S0yyLXNFkujedcfKMFEQ2YUUBd0p7m0bcaxq6 oboMX0KBbNvrxBOJJEnxcDPzMSEzriSDvyizyuqcZURxZNzIh4QQbKi5h0KGmntty/b61fGVCFa9 a9zry2wZV5Vx0uIrOiv90HqSATJgAhWHPDhkJyuQznGdaeeDEaqEkWwJmB6GsaqbFIIg1JcSMRlw AFFgHKZgNR+gK1YgALPW1DGoou051Rv4i0KMzxwNN/NKFFhdMbxtcFT5yflXv2edEU4jLzeTx+ph AkrqM4iLnlaE0ajJjTHHtorKtn7qBYP/cMS0qcQxV8TZenE5uqMaGbAC6xWXtatvig3RADqRkAH0 DdlJKAgHsAWFMJ4ntAJ6x8qb7X6ITRqNixS80wbSQt+RrTOtkN8LW0AcVJbnqn9TGr45sUtfK4Z8 KHfWZ33Q14V8/d38CZUkLneCnclNHKAFDdQeUy03ELm30hVXQ9PUTR96yqw4Hr01oAA0wOMKAAEK EORCngCcnBWzXNzg57nB2Zsa2Zu/SWC2XRTRUDi1gABRAEYk0wrXYAqKQ8CWHZFI0S3xq9kCvp6X +t7DwDszgOVRoB6tEA0BgDI/86QS3p3VTeGydUmf+8QDK9o4geFu9ovW2p/bLZIhvsn//5zo/kzE SN6mOPve2KAMbJMpbgM3LIAlO9Q4NW7jRMcVviebLdjPSTzog654RUzSfJnYkerk/vqKC+beOKkM U947QxHTMbAFxbuctgrmXJnYpgrP/EO/XZPS2DAM8U3rQtEKDnDrYZABuv4j4XY8vm7TnFOqaqmj n82jd6G3NQHoP3yUOh690cvj4z7kQy7LIE0I5d1oK04UoFMLaLJo0hDnelJA6Hxpo2lso4HVRCHc yabuxi2i7LHqTV6iTs7GsJ4Un1PrLFACL+7HFd0fsHF0sYFecLOD5/3r9RvsxbfGllTgRaEMvNCd VuCwhjsKWSPbICIIsTEUU5Lx1M5X1v/e6ihKyNsOsx6ppeUV1kjc8+Gd6NjNbow+0E77yDg5DJ4T OuqhM8L8BKRwtV+ZoMSluZ2+FdaG1cEd0gFNxFz/Bet+lYcdnNVc83Or3p3VyjgZGlXgMz3w7D8i cdPBApLHIlbQCoGQdTUUumV+SGNPoiO653B55mnfGjHA9m5vpBJHNW7OIrARvDBP22Xf9/2qls8m Ibad87i9FlYNILNMC8gGBoVN0G6aFKBDoMFTLQpU58nO3nYIQYiRgqG+obIf2NRq4vB1+8lGeYRA 9N9b2x3JFfjtRPCzQfZjpK5AKalgHVTy+AH+K6zU6tkZ+arDqu+HvHPeJ3n5I3Wf/db/UcB5j+eM i/NYmPnHg/Ugbc9b7++9vWu8f9wCqjra87GoTzwwosHGgzwTS/VyDfxZ3f/qP8QAQQubq4EFCRL8 gqUFNoYNHToM8FDiRIoVLQYAkFHjRo4dN2LMaBFbgCE9enzY8kdllT8rVL78UyUAzJdVqrBoGJGh R54aQfYECkBnxaERAxw1ipQhUqUifwYF+tNpyZMpY6pcwXJlzABaaa7AubQhVI5DRZ5FO/EpWZ5r 2b4N6hZoWoaDGipylcgV3kSK+g5K9DewIsCFEwl0RQvMjIVo2WI8G2DChw9h8tT807Umzpkrj6rU aRYuW7rYEqUVyDB1QYaJEROaMSOt/9nStSnKHS1Uo9NCKTOEYRn8z2XhwrPGCBPDStiRDnPjfmxb OtHcHaVe3BJmy5AbV70Wr1llxZYbW1ZcyekcLu3p7aGvrx7fOtzSp1nfJ4gYm/6DBg0qakE2x8iC rKJWAqCMspSE6+wqrrzDLDT15POoNEH2w2aQvFwZhDAOOSQsxC9EVGTE/r5gbLb2VlyKwo9CuijB D3rY4jKV8gjjRpqquCIKGSag4ML0xhrtPajYY7E2Ix+DkSgZaRwuyhwvs3ElFq6QIYMGhBSLyLeQ TJKuJZl0Mb4xd6PLrukEYhNDghZr7KwvnRoljzx6sIwmPff0rEsvy0QzLfvOarPQ/f9oSSxRAAWM LMzpzjyySYpYyCCDM06q7IYweqiiAe3C0PQG8dDDppUouPQTm+coBNPRyAAFySJprpjApDAoCzW5 CXLVNKuwWmHhFIgeKnK2o0ZS6tjmWlULUN2chY8+ulo5aEMNO7zWww5HHHHbEv8LMK05K5IGwQQX hBC84I5q4IMJokh1J2glPevC/vLz794S9+UWr4EUU0hFV5WEtkCLWoAiO8qGcHcCGWidIOKIryD1 NmKLZHVg2yCNC4Czonhii+0+YDjiGaKQWOIrmFPr4nHRkjCnpJoTs+B5CcytNLtWc4jnhghSjRbE hEZUsRQHjM5JGT94qcbKrPrDKp3/DqyYvZs9TlO1iYDGsK5BCvs6kS+IdmUxRp3SuDSOO0aLDwom 0LKKCfgIS5Aro6AYbZe/zDjtmmHFWqRBGsgg4hXevaIVbCjGmwVUW96bQLWVTcoommez+eqocp5W P/7yLahbfr89aFFx37pcLZv+wPSDDD44wITXX3ddWYmQvNpC0O/dfXdEsUExTr39frXgtAShmGI+ 3JNoVfmYJR4i49ESBO/kmY/8SDFDO9byyjHPXPP5OE9LzYoMpSURsL8GLFGjz7ZIWmZppSyDPDIw IQBpsIkigPwDqNjtKJK7NKEPQ2wSmpv+5b5EuaJs4YJZ9Io3PQkOKXuRel4FnTIv/+gNDHfr0eBF brY2M8WnNNTqHe9UmC9EmQ5pZLHgRK6QPxYAcGMVIeDxUrjDFSYGDAGLYAibxUEhgsl5ZhLibYgY QiOCMIm3uxoJMWbCrHHNim5KoGv+JTQGlu1ockJdDD0Yv5vprDU/y2IauZjGBibKbAJ7Yk7K2MEk teqI1aFj9KTYk9QRz45OjKMcoyi+Z1FRUD3kIQsPAgYIgpE0w9IY9HJIvVLRQkN6wVaIvNWvfoGh RGbwZPDgGMg98rGPkfxjtJAYSEE6K16RJOPLWFnKjg2SQmZEXwL5U7QuMtBowoslDB9iOxYRc4Bl pMthWkEtvfzFLyHqCyEUIU0wEP/Ck2DAZjax2cjhkRKZp3SUJKe4SlaOZF5idJU4ZelN8R2Fg2U6 oQOpdS0QZXKTo+MkNl3oSGEKsJinOyddBOGFGShEIYw5aAsSutAAKbShDB1lHGlZIZnBUiR3fE45 WwQtSI7xooCcJSEzMtGRkjRwaLHLMpnZlxApoqUlIgQ1Y6oIbWbzix99pFqMCbOd4vCbGlWSRk26 EX+GiY4YLZZQA1rRcL5QckoVqVBK6U5XJhMLWChoQxmzVa12latf1SpA+0kUOjalNJME6vaYsla2 ttWtb2XrT5nqnjwidT1wxWte9TpUei0lj2QV61PLydeoIOWuck2rU4NiLLy2B63/ieVnVCV70mH+ FSKWtetkNSstnUqnpz5dp0QlS1WpQocpUl0qZBU7l7Q+VrWg3Wxsy0LawhK2LbLFrbMMeyTatnOv vwWubT2yW5+4dbapfW0woZJY1ya3ebmFbknxulnhRneyerVudjlI3C9x10XOjax2ZQtenIrXvLGt 7nnVu172zta7vO3tLckL2/ZGdb70rW9+y5Re/fbXv9QtIXLv+6f/InbAqipwglWpYAY32Lx7tO2B n+tgQEn4ghR2MH8xvGEOlyXAHLUwgTs8mhCLeMQJ1vCJVcxgCBt4wCsmX4hXvNYFFzKqKUZxfN27 uevq+CM+HunzRNpiAR8YxiQu/zGCZ1zc0/6EqsZ9MkaAnNkjY1DKTgaJlEu6Zd1gmbQmzTKXn/ys L2v5Ke48LZNtbEsqG1LGVc7pm5ccZJ+g9st11nKXURs+OPNtz3m+MqDDTNss53m/QS40otGMaOn+ mNFh1jMhiQziEvd5rBY+spMbPWZFm5nRa3YRjguc6Eg/Wsx7pvOiARdpQrOazL0dNJk/3c4PV5Wx ebTcMF8ZXvjS2dK7fc9aRI1hTQ/a03929bFXbWkyjTnWgjZ1qg0d6k87O9lbLraeIT1tNo/T1kGM GbL86lfbsdU24zx2mrHcaOlq+rxnfjWnS/vnYf+32IZFs5fzjewpt5nZO0Z1lP9Jbe1A+3q/ZRZz lIvLaoKrWcOTxm2Tu7zuUmt7tMeFN71TLW1QZxfYqJb1q0/9bxjXm+TDfdGIcQxx2Qoa0vx+Ebc1 J2xgKxzZIVf1u2Ne6G0n/ORVNvnPU27wDPdbt7Vu+ayhXXFlSxrP7h54z0UedAwqGuccv7nQO0x1 rXedzzVGL8dbPXFji/rj+bZ5vNtNdPHCG+1RvzKXvb5hrs/d7kj3893DPvSO693vbf974NfLcsHT mu91L3zi2a54xl8X742HfORPjHjJ+53wlce85D1r1hV5O/OfpzYeQe/50Vv61mZZ63RIX3rW40z0 ra867OHslM/ervbHzLvsdX//29fv3jpG/3xwhQ9XzQF283RZve93f3nfv7f1lo2ka29vrNXyWPnX 73vsr0/j5UNVwNDvJu7JhH3lM7/52/c+pXf9T+Vqn/yyN//y0T/YHH4WWNgQhEIe1yj81vL98u+9 96O8kku/qmqVK7CCKrCCBBSP/Qu/CRm//4O/xyu/+QupOaqIKwiAFViBo+DAlkgctRE/15PACQxA 8hvAGSvACmuVAEiFlqiCF2yJzFAcEZywCCxB1os/3UtBFQO/dIoiAwmAaWAFAUiFQBgEF+iKAKhB G4RAEsxBHaTA8yu/FQw1p5iGFxiAVHgBAzCAmWDCG7qwnujBKKSwHTTB5rPC/74himmoBhdUhVUA hCVkGTEZQ94zw9JDQ9grQ5VbwxKyiBoSD5sgxHW5EsdxQiXDwTzMvD18Pgtkp6ObCLu5gqzICpuw xJhIha6wxLwJInnJPUZsxCkEQDWkv6OjrAy5RD6Jkjs5ABvJCiz4REWEQlHEPEeUwio8xVV7iEFw ia0Aj+2oAoaBHe9Yjk9MPltssOFjRuL7vWaExjQzLz/ym7lDEs7zm6G6joa4AkwEjZfAkR74A5JJ kAP4jrIyMetTRrpLK7n4QQ9SrySzRtsTi2TJNWRJPXGjGdrQxkBhCEFYgTx4gpiogssQx5PIgAM4 gASZieJggSYkClBcxHVcxv92HC6LPC959Dr2QD1yW5YuiRnvOaV+pBdaGYIhyAOURInWOYAAyIA/ GIIGmAB0AY0gSkaKHDWMLAudFC+N7DqOhKTuuRzuEUoJ4UhUpCz6IUeGKRmFDIAtIBnCIRlx1Iow 1JubxEn/ekcgtA6e1C6f1DqgHDdxy7XuKct9vEakfIhW+BGUcMseiAGkEMcPOAqWIJkqQUdadL+s RLGLwEaPtJgJ+givzC6wFLprpByRNEqmqMexHKZlS8VWuIIDWIED2IVLeAIPHMceOABsEIAemB0H scqLkMha5Mu+jMhlMTeitEfB9AnCtK40+ZBraQXatE0OqU3cNB/H2sgVyUv/KDo0f+TGo0iCWoiE EiiBp7yVd8EGFpiMGbER8bBJsDtNeyPNj0xM7PSeDnJH2IyuaYFI20gEL3BA5OtN7PFLJTKTl0tF bPCCACjOTggAcxyCSpkbbJAGBNiClyQOmTg94KrOiiQr1szH1QQn6elKoOrD9sQGQvgCB4XQLACD LwCDGgADCa2BL8hQGvgCGqgBirALV/AC3vxJqIrGp5iIPSiBWugEpmGYDJABE4gCaWAFc4kSKxGY AGW2rQyni4yowPybaZyIbLKmCZ3QL0BSJM0CDc2CGmjSGqABimgFgZBF1TvPy+opfnSP58m5VIyG YTDOSqmUA3hI/DuALSiE/5c0jvKEHB39Ndj8q6Iwy3wM0p6ciBLZi/TZiw0hhNmsJ0VwgQ+diBDF BoWSjnnUNXKb08QsUMYcot5DUYf40lqoBQSIAhbAVGmQhmhghYWkjBttiSo1EEhy0z7jUaPy0U8U ScCs068c0j6tDUKAUoqwDwdyBfipvn8Ty6EEycbs1XBD0OCkLGUg1lpYSxlAju2gyqtYATYtlQMh 1VIFOsbCx5n5SLVJ1ciY02r9yw3KyIlwUJ+ZiNXYCw9lCEXQU3S1l1IZCFwtr8NMVEWtHKHkVpnp VpJsT2XghYYAlgC4AcrogSrhiodswtpcJobQwI7US2n1Q4FBvZmBvu5U0P94HFIw4JryoYEoHQi8 4NjDOBQHAiZeIzktbUzu2cdxC0kRWrbIEBnWOQPiIMgN3MAVYIgD6UAraAWA3MRoZVgVVKt6LEqC SdA/zC2KIIQscIURyQJZBQO7cIEsoAUNvVANVQQFEFS9OCNB0NPD0AtDDayf48emENux/R55VU9h FQlBOADaqRGYkInveMgGkYlfLIp07Fli2x6PVMzpC9bBnNhvlQgKJQxs8YIR7VAa4BAHrYGvkVUF aA1F4NdBwAIvuKqrio3Y+NqTO1XfzJypKIkP6A7jSJfMcJAosQnmGIq79VkrZBZHLSrI8VYhlYgJ TVpC6BAs+IIGVQAaaIX/waWBfbHac02gf6RcgiIogwrZEQRbop0ip9AOqMwTVtwRm9COY7Qg1fVB OL3O73krfNRHlZXdh0BSoDnagfgaGnCBDaGBBKiLjH1cRDmIZnqmZnLXiUDUC+RFIXySQoDZKtCO 0T0PFvgRHEAV0cDehmXd7cVOoF1gvl1QiqhQoEFSDZHQBKCBQahQGpDFvdhdhmgma+GQWs0QV0he +73SSAQckXBOk8CUTwGOWgmDT9kC8fgVROyohT1gAU3gAdVbt/LeezxbOw1cT6oLLPgaMHjaJfVQ 9kUIGnDcjWUplvoW1XCp+r1BeN3FFFZhK0DT1zkJuJEBLCAcuAESvLlO/7vNYdS84WtEz43ayb8N X4e4UAlGUkWAUg7NgiyggRlQHFcghA7GBnTtltn0l7og4cwdWebNqHpJGbnpEYbgA+u5guUJvzTm sIdlzHklS3skyh82JolV5NiiiCWFXOChUCgtmzx2ARoYUYb4Agh44j3lC01yCL+w4jt80yyGlbTg gx65ghZw1ge0ZHZMj0wmy01eYJRN5lby21DeLIrQUKCx0BbgXWyw0ARoAQT4jwSAAFcWmxL55i8Q YQ0p4SteXl3e5SSymmE+Qyw1q6LU5LIMWmwEZXSOzYkgZVe+Y4aoAShFgATgmkGAZQ9mH7EJG/Oh hSpGZF11ZvrY3B/FYf921krtpb1tJdsDDWJXDdwaSAxs2F1BqEGrRYCcLYjaBGSJIFcM6ZByxmVT begweugHLE2Jzkk4bubUDDeFtawHnggnLWW0EIgvUAAFyIv5fSbAGJRAHoRbfsJzxl9Ki2kgNWea nmibfk29kVPH7FbqAFzx/d0zOiPe6WNCSIAnXo2U1o+VXugdfek4i+p4NWGqVmN75s5yM1mU5Vs3 FuKHcNLd/Jm/PiOhZl+LIFeFzlW2tufvuuGmaj+51q84hevLatWR8k7oguYa0KR92RYHLZFBeNAv GISyvqR6+tMTAZC1zuWn/jbHLKYOcuy55mFO3tvJForKLtqJ8NDcvuP/jIUAJ4ZlWB5qBSjrwRYJ Npkmpkbj1Ebh1UYnJamr17bORuHeY7ZWmMlW1S7MiehQKK0Bq/VuJ+5uCLDg4RZvBWiTsAadEWHp ptbctr40yd4YzILuqq5o7/Xk6nZNyrbq7EYpNbFVlipqvhgMpB7cwviZmELumdbcE4XGlgM+55vv S6ZoqQaTFszve6Ye+zhrwMYPBeId9UbtCGcyZxRxAtzv2rY9eObWeV1U+wbOOLYI9WmND+6LGj/q AVefHN+ZjV2MEC/xHy/RE69ws1RNiK3WI//ex+xqwZlS09gQTCLkbLGnfTHt9U5uIMdyp6ZrnaKc Ik+WI8/kz+JptJBx/0JpiJRujSxCcB/P8jZ36dRcPwuP3ZuG87QZ83pp8tHW88z2lj73F4Q47cN2 80GHM8RU8dVMvTCv8Otebv4+izIH6gOSdAaypgSPaELHdBNPceom8nge24zWb/cWKYFCoc/BF0Ra IRAX9ExndQlPcaNccS8HczFndNFacov4mofYcLDejxwH7cIgG2yy9FYn9kybxegb2sS27OPBWlRP pBMhBCtX8GKn9gybcGW/bTLfcQMqlCxaIF5CFGFn82on9/p6a8+q9Se687RFIWd393/xJGm/9HKn d3O/duzG8LPIdStSoy3aInC31XDv8VWv94J/sHtv9HxPW72wJD+dcv9OGp1QsiYz+AIzkHeDx/jH ZvBoHK6Nb/Bbr4hcZ6ZWcKaSn6aTr6aUL9KaGvaMd/mXf++KEIREuALkRaiHwvmb13mHuvlxh/mf p3fDBPqhd3mhJ/qjr3ejR/qlt7waEzb/Unqml/qNXDhswzeHY8/6ivqp53rN3biXAzR2yz6Qv6+u N3uql7UzU7VYW7z12vqzh/uS+/pT27aw76+3j/uXX5IzwcW2f4x9s/pFC7Q7u/tKy3upjzurLy2e izszwzd5czaeW/w6+zy8P/yMT3xOkzd6ozlTK7OeSzfQs/zLN/jEx7m6R7i1Iztfs3tPy3rNM3zS R/qzKyS2V7ilM7T/3L82uz9B7dP9wpcz2Sd63FezyRd84799tXc4V6u15Id84wJ+TBP+w58oEhIu qAP7jZs4v5/+7r/F9eQYCA89kSt+0H9970d/Vsf+kTN/7k//98fymgt8uFv8BYX/+8f//Nf//ef/ /vf//wcIAAIHEixo8CDChAoXMmzo8CHEiBInUqxo8SLGjBozBui48SPIkCJHkixp8iTKlCpXsmzp 8uVFjwA6yvQYQCDNnDdr4oTp8yfQoEKHEi1q9CjSjTd7zpy5s+nSpVCnSk1q9SrWrFq3cu3qFaLU qFSZ6nw69SvatGrXsm3rNm3Ynk930p3b1OnbvHr38u3r9y9CmnjHlNJ1ancw4MSKFzNu7Nhn1YSR H1OubPkyZsqTDcrM7Pkz6NCiR5Mubfo06rc6wXKW3Do17NiyZ6+NGtZsTsO2O+Ota5g28ODCh0O+ W/gw7oFijfO8S/w59OjSJ+62u7wsztzGt2+e7v07+OHL5UKtaxYxWe7h17Nvn3o3+bnmlQvWbTu7 +/z695vuzv8/gAHK5h90AQEAOw== ------=_NextPart_01C52880.DC295B70 Content-Location: file:///C:/C9344E54/Paper14_files/header.htm Content-Transfer-Encoding: quoted-printable Content-Type: text/html; charset="us-ascii"

- 1 -

------=_NextPart_01C52880.DC295B70 Content-Location: file:///C:/C9344E54/Paper14_files/oledata.mso Content-Transfer-Encoding: base64 Content-Type: application/x-mso 0M8R4KGxGuEAAAAAAAAAAAAAAAAAAAAAPgADAP7/CQAGAAAAAAAAAAAAAAACAAAAAQAAAAAAAAAA EAAA/v///wAAAAD+////AAAAAAAAAABoAAAA//////////////////////////////////////// //////////////////////////////////////////////////////////////////////////// //////////////////////////////////////////////////////////////////////////// //////////////////////////////////////////////////////////////////////////// //////////////////////////////////////////////////////////////////////////// //////////////////////////////////////////////////////////////////////////// //////////////////////////////////////////////////////////////////////////// ///////////////////////////////////////////////////////////////////////////9 /////v///wMAAAAEAAAABQAAAAYAAAAHAAAACAAAAAkAAAAKAAAACwAAAAwAAAANAAAADgAAAA8A AAAQAAAAEQAAABIAAAATAAAAFAAAABUAAAAWAAAAFwAAABgAAAAZAAAAGgAAABsAAAAcAAAAHQAA AB4AAAAfAAAAIAAAACEAAAAiAAAAIwAAACQAAAAlAAAAJgAAACcAAAAoAAAAKQAAACoAAAArAAAA LAAAAC0AAAAuAAAALwAAADAAAAAxAAAAMgAAADMAAAA0AAAANQAAADYAAAA3AAAAOAAAADkAAAA6 AAAAOwAAADwAAAA9AAAAPgAAAD8AAABAAAAAQQAAAEIAAABDAAAARAAAAEUAAABGAAAARwAAAEgA AABJAAAASgAAAEsAAABMAAAATQAAAE4AAABPAAAAUAAAAFEAAABSAAAAUwAAAFQAAABVAAAAVgAA AFcAAABYAAAAWQAAAFoAAABbAAAAXAAAAF0AAABeAAAAXwAAAGAAAABhAAAAYgAAAGMAAABkAAAA ZQAAAGYAAABnAAAA/v////3///9qAAAAawAAAGwAAABtAAAAbgAAAG8AAABwAAAAcQAAAHIAAABz AAAAdAAAAHUAAAB2AAAAdwAAAHgAAAB5AAAAegAAAHsAAAB8AAAAfQAAAH4AAAB/AAAAgAAAAFIA bwBvAHQAIABFAG4AdAByAHkAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA AAAAAAAWAAUA//////////8BAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAANA/Kom7KMUB /v///wAAAAAAAAAAXwAxADEANwAyADMAMAAxADMAOQAyAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA AAAAAAAAAAAAAAAAAAAAAAAAABgAAgH/////AgAAAP////8AAAAAAAAAAAAAAAAAAAAAAAAAAAAA AAAAAAAAAAAAAAAAAAACAAAA8csAAAAAAABfADEAMQA3ADIAMwAwADEAMwA5ADMAAAAAAAAAAAAA AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAGAACAP///////////////wAAAAAAAAAA AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAGkAAADy3wAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA//////// ////////AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAANoD AHic7F0JXBNHH50km3ATbvEMeGvVgPfVLiqeVaMI1lrbiKAWr6DgUVGDStVabLSitS01rS3aWjFq xVorDVrFo5Z41R5a4o2KdfEE67Hff3NgCCG7QLD4/eZ1XrPZzM68eTv/mdlsZI/rvM9/sb3eBWSF V5AAPaVdkMhiHwHkmd94IcQ3vX9K07RlHiEwDj6gMWo9ngDf5hnPYSG8PjHtc4Jz6Ax0AboC3YDu QA+gJ1Bs7ALIG+gD9AX6Af2BAcA6wEBgXWA9YH1gA2BDYCOgBBgEDAY2BjYBNgU2AzYHtgC2BLYC tga+BGwDbAtsB5QCQ4ChwPbADsCOwE7AzsAuwK7AbsDuwB7AnsCXDX0bIRIYBuwF7A3sAwwH9gX2 A/YHDgAOBA4CvgocDBwCHAqUAYcBhwMjgCOAkcAo4Ejga8BRwNeBo4FvAMcA3wS+BZQDxwKjgeOA McBY4HjgBOBE4NtMHAEnAScDpwCnAqcBFcB44HTgDGACMBE4EzgLOBs4B/gOcC4wCTgPOB+4AKgE Jhu0K+C/RDgXfaHcRCiLOYI7AiDazX2JiX3fenzD/hzjx/0s84b9tm7Zjy0P8gTMG4FxHw+cnGJo ddXgivg8y/7M9Tg/06sA2j0MXBgIZzWyCvV7wSjIQ8bmcK2fGVM/32/0iQc9TwFnNR58GGc405WD TxXaz+idbtpm4p5vejWT0Wcr/p/W7FCE8R8A+gLf2ap/rBxZl3P/40GvELga+7517DPj/5C4mBmK BMWExKCRcQlxiqDwGdGz46ZNNMwbIweOGCgL6twuJGjE29Hx4xMME4ohWztTtnad0b1uO6ZXULcR AsOcVFU8pZk1C9/hxzN7+L/9+lt6u/peqz9yRi+1ebiVmV+EVvtuwr6vkTHemLgMMx6ezMw9zDzL zDHMXDvGVCYzJzDzMDPWM+1OM+2/TBjnYBfCOL+a12kVbbub1m8LgYuAi4EpwHeBS4BLgcuA7wGX A98HpgJXAD8AqoArgauAHwJXm7SsAa4FfgRcB/wY+AnwU2A68DPgeqAa+DnwC+AG4JfAr4AZwI3A TSZfvgFuBn4L3ALMBG4FaoDbgNuBO4DfAXcCs4C7gN8DdwN/AO4B/gjcC8wG/gTUImM/3QfcD/wZ eAB4EJgLPAQ8DDwCPAr8BXgM+CswD6gDHgeeAJ4EngKeBv4GPAP8HfgH8E/gX8CzwHPAv4H5QD3w PJBZeF8EXmLOIfAK8CqwAHgNeB14A1gIZPrKP8BbQApYBLwNvAO8C7wHvA98ACwGlgAfAv8FPgI+ Bj4BPgUaJgroDDxmAAAKgARQCBQBnYDOQBegK9AN6A70AHoCxUAvoDfQB+gL9AP6AwOAdYCBwLrA esD6wAbAhsBGQAkwCBgMbAxsAmwKbAZsDmwBbAlsBWwNfAnYBtgW2A4oBYYAQ4HtgR2AHYGdgJ2B XYBdgd2A3YE9gD2BLwNfAZLAMGAvYG9gH2A4sC+wH7A/cABwIHAQ8FXgYOAQ4FCgDDgMOBwYARwB jARGAUcCXwOOAr4OHA18AzgG+CbwLaAcOBYYDRwHjAHGAscDJwAnmtbkN0xr8snwOgU4FTgNqADG A6cDZwATgInAmcBZwNnAOcB3gHOBScB5wPnABUAlMJlnHA+YuHYzsRcM/QOYAcaEo7DwbA4L6qC+ Q/oxuZl1SzO+cW3AjEtTTPly4CIgJ9AYJ0w/sUROXQHKqUPAunggrJMHGtbO4bDSjIZVahysOiea 8jFrf6ZqKDvZvA3XksnM+jrepDWdLrvGWfmsrqBeUGYclDoF1Tbst7mXbd4atf2sYbxnED/7nIH3 in1QlKRkdmCzktnMnEuU+KD2/sxS31jHWOA8yMd8NgWOd9rVcSzzyiAImArbTLmnszqOZfKNMn3G IHaW4bSWngeexTbMRcktLfYza0IvC60upvezof/84m68hmuAns03lt8jWJbvY2ObmZeY679A0zYD 82lvZpGHuc5irgGZfvsmzHujYOe3IuP3D2K0SRTJZxhNiBHDTSLL/mXZFsu+1tKi/LqmbaZucx9k 9tvqg6Z2OqAPnkPEHR+0TLQf7b1jPNdhrY3n+hi8b3To8exAp/3o8AfnUJrqrOHVgE09xjLXIN/4 oTlMXubVuuRrK84ZPGXAHMvQXIe5PzFnwdyfBBZ5zWXyIH9W1lnUFV6Z8oi7xuNbWhwPC43S/mit 07zd8a6P4X1dnyZjGY+Pw/tJ2WiOdR9k/GaL/+ftPdNuph3MqwHORq/mLD6HwiAf8wVFyD3j8cdN xzNtZrwvgbyO8J4pj817a53mba7eW8YGjFXJAabtQKvzE2jaZmAZz8z1NNMJme8+3CC4nCDzXKHx s6iopWELRy8NYz6bIDSufwdbnQ3zNQqz9jVfJUfDfMF8Q8Gsj5nvd0aYXs1sj8qOI9Z6mO9AmO9w FkOd/pDhjpUe5rN8oXF+s6cnHPTEoJmgabzhe5MBFvVb62pgpcnSU/M4aT7ftsZJAWIfJ4lnGssc y+y3HCcvOBnHSamreZx8yTWSz/CKkxgxfMnV1jgpQLbHSaZ8W+Mks782xOrKN4wxwLyaIbBTKgNz rDJ4XrFqrdO8zTVWGb/N3jN6aoP35eYo0zjpZWecZNrsyHGSKa+m5yjL2KhonCQQt3FS5WwcJ586 lx8nbzg/33GS+d7aycWoZ5ZLWT3MZ+NcuIyT0TA6RqNxwAQYK816GtjQ0xVVPE5ajknBxlfD9xmM p3/ACZCBwV1g8NwjMt6DMJ8Xy7HJVtnIomzLesDnZMvzFWCxzXznHzv9Wb2doM7domfvzYhGxu/1 me9wWprO8c1442dB6Bh6C4RM4hs/M+flmUYeJt8BTyOtfcWoENgjdmCP2IE9Ygf2iB3YI3Zgj9iB PWIH9ogd2CN2YI/YgT1iB/aIHdgjdmCP2IE9Ygf2iB3YI3Zgj9hRUx4hNtRQvTUB7BE7HK/V7IGt lJLCJNr068CKLKtt/mKPOLTIYQWVwmiEZbLpl1Uq40FlMte8ZdgjDk1zQBEcG0bbss+mm6zZrJIt pxxoIvaIQwOreTzNxaPKtrlqyToCrWMTe/RCeOTA2KlhE7FHtdmjasaaHaMrbm11zMIe1TKPyjhl 06zKhgPn5tnLbwns0QvoUXm5lTWrYrBUXUFOa2HYo//Yo9L4sZvHUmk1Qs+uL1UAF7ttH1jJarBH tot+ZgoXj2zIqij0KrMs4N4kjqoqkZ9DidgjruVaJRbYEVfFAHS0U9yBPWJHFcdsU7Jz/VGxX/YC sFY6hT3i0OrK5bZouLVTNodkW3Fpu8DKx91zMwt7xKHVlctdrr0VOlGRX6UDt82SOcYd9uj/zKPy TtnKYzMMyzbSdknczHK0J+VaXbnc2CP23Ngje/ms2lB+4LD6yI6ZlvkrcrN8gdijF9sje1HAmmzC 5pT3Qsca9qiqHrHGGqtlVu0v75Qda7BHL6xHXBSXH3HthJWd4OJeI/boxfQI2WoAa5O4hGGlPCp1 +LkAe8SOGou1qoUh9uj/1iNTPoc7ZdM4bu7UqCM22s41H/aIa24rONAjOwVaoIZcYGl15XJjjyp9 sB2Uax6LsxXDUU2tejOrdTD2iGsR5qaUSTZ3Wn5ky5rqi6kJYI/Y4XiPyly3WuyxvqQt61H1ZdQc sEfswB6xA3vEjtoyZldfRs0Be8QOx3nEHbaWjtWXUXPAHrHDYeLsWVBrvuWoGrBH7HC8SsuoKr/n hQguK2CP2PEiaf2vgD1iB/aIHdgjdmCP2IE9Ygf2iB3YI3Zgj9iBPWIH9ogd2CN2YI/YgT1iB/aI HagxKn1uI2357Efz83KFpu97bD0vl/nM8nmWrwEJ2FAJEHqVQMgfDjoO7A77vUUrhGLRJKGLaICQ FjYX3hEKhReEV4lfhUeJXUINkS78iEgWLiLGCacTvYUxRJBwFPGUiCDOEjLie2IYsYaIJBKIN2FP HNGeeIfwJ1KJfwVfEhcEe4ljgj+IHwQPiM0Cf6Fa0EX4keB14WrBPKAatvfDvsvCzQKh6AdBM9Ex QR/RBYFc9K9gjsif+EDUnvhCFEFsFyUQP4nWELmi74lfRGeJY6KnxBFRkHC/qLdwt2iccLMoWbhO lC5cLNolnCL6VThcdEHYUXRH6C+ihfehTaegbduETBttPdNXWNazKnvbFBjDR2g2fPgWeJwsND5j cyI/STga2sulblvPNBaismhm41imTOb5xm2Aw6HuN+DD1fD6mdD0TFzBG+D3fOFwwWdCZtvyOFvP M2bKtnxmqL3njuZDPTlOCA0Co9q4Pr/njprrHQD1tnZ99t4MtueOTgYhk/FzRx0J7BE7sEfswB6x A3vEDuwRO7BH7MAesQN7xA7sETuwR+zAHrEDe8QO7BE7sEfswB6xA3vEDuwRO7BH7KgpjxAbaqje mgD2iB019vttm39oJKXcn4ywZVlt8xd7xKFFDiuoFOX/6QjrH0SysqVSmWveMuwRh6Y5oAiODaM5 /P0Sjtk4/BMnB5qIPeLQwGoeT3PxqLJtrlqyjkDr2MQevRAeOTB2athE7FFt9qiasWbH6IpbWx2z sEe1zKMyTtk0q7LhwLl59vJbAnv0AnpUXm5lzaoYLFVXkNNaGPboP/aoNH7s5rFUWo3Qs+tLFcDF btsHVrIa7JHtop+ZwsUjG7IqCr3KLAu4N4mjqkrk51Ai9ohruVaJBXbEVTEAHe0Ud2CP2FHFMduU 7Fx/VOyXvQCslU5hjzi0unK5LRpu7ZTNIdlWXNousPJx99zMwh5xaHXlcpdrb4VOVORX6cBts2SO cYc9+j/zqLxTtvLYDMOyjbRdEjezHO1JuVZXLjf2iD039shePqs2lB84rD6yY6Zl/orcLF8g9ujF 9sheFLAmm7A55b3QsYY9qqpHrLHGaplV+8s7Zcca7NEL6xEXxeVHXDthZSe4uNeIPXoxPUK2GsDa JC5hWCmPSh1+LsAesaPGYq1qYYg9+r/1yJTP4U7ZNI6bOzXqiI22c82HPeKa2woO9MhOgRaoIRdY Wl253NijSh9sB+Wax+JsxXBUU6vezGodjD3iWoS5KWWSzZ2WH9mypvpiagLYI3Y43qMy160We6wv act6VH0ZNQfsETuwR+zAHrGjtozZ1ZdRc8AescNxHnGHraVj9WXUHLBH7HCYOHsW1JpvOaoG7BE7 HK/SMqrK73khgssK2CN2vEha/ytgj9iBPWIH9ogd2CN2YI/YgT1iB/aIHdgjdmCP2IE9Ygf2iB3Y I3Zgj9iBPWJHjTx3dIMAob1OCLWAg3q5Gp872kREOweLzjk3EP3g7CP6yFkomu18TzjK+bywh/MR YZCzRujivFp4zylReNEpUnjaqYPwqJOX8Genf4gcp6PEPqdNRK7Tu0SeUyxx1qkvUejUknjq5EH4 Od8TtHXWCwY6HxNMdM4WLHXeIfjaebPgqPNGwT9AL5fNgg4uOwRRLtmC2S7HBGtd9IKdLvcEx108 iOsuLQnk2pfwdY0lmri+S0hdNxFdXY8SPV3/IV529RJ2d+0g7OgaKWzlmiis77pa6OqqEd53OSL8 2+W88GeXe8KNLkLRUhcf0QSXBqJwl2BRYxemjVye/WnvGZuH3JDhnHTwRGgF//k9Y9NcbyjU+z7/ 2XszuDxjcxJ+xqYjgT1iB/aIHdgjdmCP2IE9Ygf2iB3YI3Zgj9iBPWIH9ogd2CN2YI/YgT1iB/aI HdgjdmCP2IE9Ygf2iB015ZHN35VaoobqrQlgj9hRY79VtvlHNVLK/XkEW5bVNn+xRxxa5LCCSlH+ n0mw/vEfK1sqlbnmLcMecWiaA4rg2DCaw9/q4JiNwz/ncaCJ2CMODazm8TQXjyrb5qol6wi0jk3s 0QvhkQNjp4ZNxB7VZo+qGWt2jK64tdUxC3tUyzwq45RNsyobDpybZy+/JbBHL6BH5eVW1qyKwVJ1 BTmthWGP/mOPSuPHbh5LpdUIPbu+VAFc7LZ9YCWrwR7ZLvqZKVw8siGrotCrzLKAe5M4qqpEfg4l Yo+4lmuVWGBHXBUD0NFOcQf2iB1VHLNNyc71R8V+2QvAWukU9ohDqyuX26Lh1k7ZHJJtxaXtAisf d8/NLOwRh1ZXLne59lboREV+lQ7cNkvmGHfYo/8zj8o7ZSuPzTAs20jbJXEzy9GelGt15XJjj9hz Y4/s5bNqQ/mBw+ojO2Za5q/IzfIFYo9ebI/sRQFrsgmbU94LHWvYo6p6xBprrJZZtb+8U3aswR69 sB5xUVx+xLUTVnaCi3uN2KMX0yNkqwGsTeIShpXyqNTh5wLsETtqLNaqFobYo/9bj0z5HO6UTeO4 uVOjjthoO9d82COuua3gQI/sFGiBGnKBpdWVy409qvTBdlCueSzOVgxHNbXqzazWwdgjrkWYm1Im 2dxp+ZEta6ovpiaAPWKH4z0qc91qscf6krasR9WXUXPAHrEDe8QO7BE7asuYXX0ZNQfsETsc5xF3 2Fo6Vl9GzQF7xA6HibNnQa35lqNqwB6xw/EqLaOq/J4XIrisgD1ix4uk9b8C9ogd2CN2YI/YgT1i B/aIHdgjdmCP2IE9Ygf2iB3YI3Zgj9iBPWIH9ogd2CN21MgzNg+4IQQJDfREaCXf+IzNvp5DeX08 X+K97OnM6+h5FTX3zEUBnpsQ33M5KvRQoBMeo9BOjz5orUcomuXRBEV61EXtPXyQ2EOMbrp7oSPu /miTe0O0xL01mujeHQ10H4LaucegAPd5IGAduuW2C+ndzqBTbsXomFtd3mG3nkA575jbIt4pt694 erdc3i23KzzkLuAHuEv47dy78ge6y/gT3cfxl7jP4G9yX8A/4r6Uf9P9fb7YI5Xf3uM9fqTHYv4s jzn8tR6T+Ds9Xuef8AjnF3q04fM9ffgBnvd5zT3PQFuyoE2roW3TeUwbK/uMzQh4jecZ86TTZb9s W8kr3QzqhWagOBSNptj8No4LziHijg9aJtqP9sJrlKRkdljrktnMeT8G7xsdejw70Gk/GnvxHEpT nTW8miFgKXnlG+dMT9NEhmMZmusIbGasg+kp7f2ZR3PuLy2Pycd89o0fmsOD/E/unEVd4ZUpj7hr PL6lxfGokfF4BtY6zdsd7/oY3tf1aTKW8fg4vJ+UjebEzjI8yrRM37V8Hqn5uaOBVnksn01q2dcj mXYDf2ca44TQL5BZ4WT8LCpqadjC0UvDmM/ecDLGwmArz8wBGGsoawac2WkoAc5uNEpECnifAPtf BjLPYO1qYntkfD4qs93A9Bps2m+pzdZzVm09z9UFPpARCM2EON0jen7PczXXmwD17hY9e28Gfp7r cwf2iB3YI3Zgj9iBPWIH9ogd2CN2YI/YgT1iB/aIHdgjdmCP2IE9Ygf2iB3YI3Zgj9iBPWIH9ogd 2CN21JRHiA01VG9NAHvEjhr7XbzNP+CSUu5PcdiyrLb5iz3i0CKHFVSK8v8kh/UPTVnZUqnMNW8Z 9ohD0xxQBMeG0Rz+LgzHbBz+6ZgDTcQecWhgNY+nuXhU2TZXLVlHoHVsYo9eCI8cGDs1bCL2qDZ7 VM1Ys2N0xa2tjlnYo1rmURmnbJpV2XDg3Dx7+S2BPXoBPSovt7JmVQyWqivIaS0Me/Qfe1QaP3bz WCqtRujZ9aUK4GK37QMrWQ32yHbRz0zh4pENWRWFXmWWBdybxFFVJfJzKBF7xLVcq8QCO+KqGICO doo7sEfsqOKYbUp2rj8q9steANZKp7BHHFpdudwWDbd2yuaQbCsubRdY+bh7bmZhjzi0unK5y7W3 Qicq8qt04LZZMse4wx79n3lU3ilbeWyGYdlG2i6Jm1mO9qRcqyuXG3vEnht7ZC+fVRvKDxxWH9kx 0zJ/RW6WLxB79GJ7ZC8KWJNN2JzyXuhYwx5V1SPWWGO1zKr95Z2yYw326IX1iIvi8iOunbCyE1zc a8QevZgeIVsNYG0SlzCslEelDj8XYI/YUWOxVrUwxB7933pkyudwp2wax82dGnXERtu55sMecc1t BQd6ZKdAC9SQCyytrlxu7FGlD7aDcs1jcbZiOKqpVW9mtQ7GHnEtwtyUMsnmTsuPbFlTfTE1AewR OxzvUZnrVos91pe0ZT2qvoyaA/aIHdgjdmCP2FFbxuzqy6g5YI/Y4TiPuMPW0rH6MmoO2CN2OEyc PQtqzbccVQP2iB2OV2kZVeX3vBDBZQXsETteJK3/FbBH7MAesQN7xA7sETuwR+zAHrEDe8QO7BE7 sEfswB6xA3vEDuwRO7BH7MAesYP1ea7m57jaep4rs6+i57m+Shif53pcZH6e6wphH89Jwpc9Bwg7 ejYXNvcUCgM8rxJ8z6NEoYeGOOHxEbHTYxGx1mM6Mcsjhoj0GEW094ggxB4y4qb7MOKIeySxyf1N Yol7HDHR/R1ioHsq0c79SyLAfS+B3P8gbrk9IPRu/sJTbl2Ex9xeFx52mwdUw/Z+2HdZqHcTim65 NRMh9z6iAHe5qJ37HNFA9w9EE92/EC1x3y7a5P6T6Ih7ruim+y8isccxUXuPI6JIj/2iWR67RWs9 Not2eqwTnfBYLCr0mCLiew4XBXh2FDX39Bd19LwPbToFbdsmZNpo63muIlT2ea5V9bYp0AM+mA2F eEAFSqHxGZveoiShi3MSp7otn83ZAhmf62lZv7UGyzJHwmsbRg+8GQ2F9YHXdHglDPtGC/uABp5z upDZtjzOsn5LXZbPDK2OJxL44BFUVN8JIYGLsW1NRLRzHSfauabPRwMn4/loalF3ENQb5MTtfFSn 7plQGQF1xgsQ8jTV/Q7f2WWKwNXlefWFWKj7OuzYCK8vuRj7QqzguvNG0BAreMmF2X4efcFWW5xM x9Cm77AtnjmLGP0hQBJEvQmi1a4IfSsyPsO3j/Mm0aeuDKOJPs4MN4m4+NnSonzL9pif2czsz6TL fqH+Xzyz+doK47OQmVczntczm/+uf47zM5utdZq3uT6zmfHb7D1zrmqD90y7mXYwrwY4G7269vY5 FAb5UDvol/eMxx83Hc+0mfG+BPI6wnumPDbvrXWat7l6bxkbTqjsM6bNccLEqq3naIsQ+3O0V0Bt tyBuh0HcDhKWfY4281k3IftztIfB+WaepD3R8P949DZKMIxpXZDhNBjGoHY2thm2BtZBZccda53M c6gLQEsEaHkAY+NWK53MZ58Jjf5UpHOMoaxoNA565niTvg4WOnwqoWce8E+ocxczZ4NvPqKyepjP eCLjM7/t6RkGkRKDEtFM8I3R1MVUl1mDpV+trTRZ9gPzOM831WFrnLduh3nu+wj0z4J2fAgnOdm0 DvrUda7wA1j32RqrLeuuaO7jo7JoZuNYy7nvXaj7DSjssJvxPDJz3xK3Nwxrz3fdPhMy22xzH1O2 5Vxhmd88djF50mvB2LVujnFMYF7NeF7zxpM7ZznPG9Y6zdtcxy7Gb/P54aGyfdbWeMVH9scrJu7G mAZCDRSicCobd8xnbzixx91ANA2ibTyc7UTD/5nY6Qpsb3oNRqh0XzCqOO4s+1iw8RXsM7YBiRHq CR25Oxw8AjT5m/Iy58Wyn9oqG1mUbVkPFJNs6U+AxfZw4LLAZ/V2hYKHOz17bwYzljAxN8x0bpix 1cvUsYLQMdQNBAYIjJ+Z8/JMPY/Jd8DTSGtf/3tgLbaBtdgG1mIbWIttvGBasjPSUhNGJ8QMmhzR LWZQaHWSI7TQVH5+Xm5+blZ+dmZ+dkZBVnpBVpo5pUKiMlNMKSPJmIozEpiUPrk4PcaYHKalwCLl l0kFpelMmZSfZ5kcpiW9rJYKdNlV9Dy0lBNVXpFJy2RHaSmmrBNFlVd0JjOVSRkppalUjqO0ZNrS UkaXHY9M7kA/dqiWgtzMgtyMgmwm5Wen52cZU5oxnclMM/licifV4ItJkSO1FFeQ7Hv0rAdBlDtC Sy5ooZmaDb5kVuBOuk13SuU4Vkv5ZN+msu44WguVl1WQm1Upd0rlwJBcJS2ZMaGZoyG1gpQREQwp fVAgpDRIvb1SezundDOkUJRkSLacsnanqlqyJ/emMtOMqSAztSCDSfkZKfnpKWfSkyDlpSUYU25a Amh55k62tTulIxHMWVXWkpVm3RGo0lRQbEwFBSAnJbS0BCt3yoyKVdcyiMoqH8eWciijHEZLN2fr vmPhTqkimNGrpCU3waCFLs6MCE4d1Co3O9MoB95GdPPKgq5slgNaUrs5lz3atjuwzKiyluwMZiNp dH5Cb5BDgQsQRjHdYBvkwGrGoIWCvgNd2XhUXkZK+VQqp6pa8pIiig1a8lImF6QlmOScyctPijBu p6cmGK0xaPEy+ZGXCeeIymX+b1hjpZfRkl01LWeSRhdnM1//n0lNyEqKOZOeYpSQO3kQqMuK6VZq DWhJG/RMSz4EHajITi+rpbg6WlJGF+cyWvLTkkBLQXYmSAA5Kb2DC2AQM+jKSE0CLRDoMO6UaqFy Tb4YRFlpyaiSlvyUmOLcLGYjIxW0QOzCBqhIHRRcAKve7EyjNdCDGC0RwaVaigvOwKm0mLkgrJ6l qmlJZbSAKQXMWJeWmRRjMGty2qBWoAX25BrOFJw7GAMzLLQYIogGOTC1w6WDMcFaofQaovJaClIn F+dlgxboL5DyEiLyDd0nPaIbU2VCRF7SaEjQfcppYWYtcAe0UGeyqbzsgjzD/JWdySqkAi1UGqOl 4mOeDYAwQ2SMblWqpTQx06lhSizOP1N8JpeLkAq1JNjV8kwOTFiZZi2WsNTCUYg9Lbkgx5Ryswwp EwIdxh0YBmFUNqQ0gxYbVeVmZRi1TGZOK1dUoMWcJkPfKUiNgd4MwQWxfobpKREwMEOCKRQSLDDK l5Bn0AIdl7uQCrRY4Ux2VqWKNGphxuZKgpsWi5SZV1HKMqYMY6qsEG5anhuwFtuoXVrYfgdnvt1m 6x6l5e9PLO9Rkp4IxQoR6iZGaJrpHmUfz7eFncSTbN6jtLzXVNE9Sh4qi2Y2jrW8R9kW6h4IhSng 9X3TPcq24oFCBWhoK35fyGyz3aNkyq7oHqX1/SNvX+N9m/n+CH0tej73j2KnP6s3CSrcKHr23oxo YAwqe//oZrzxM+b+0WQQwrClRV7z/SMmX+29f1RrgT1iB/aIHdgjdmCP2IE9Ygf2iB3YI3Zgj9iB PWIH9ogd2CN2YI/YgT1iB/aIHdgjdmCP2IE9Ygf2iB015RFiQw3VWxPAHrGjxv7Wps2HQqWUe7yP Lctqm7/YIw4tclhBpSj/Z35t+mX7L91X7G9FmWveMuwRh6Y5oAiODaM5PGuKYzYOf47agSZijzg0 sJrH01w8qmybq5asI9A6NrFHL4RHDoydGjYRe1SbPapmrNkxuuLWVscs7FEt86iMUzbNqmw4cG6e vfyWwB69gB6Vl1tZsyoGS9UV5LQWhj36jz0qjR+7eSyVViP07PpSBXCx2/aBlawGe2S76GemcPHI hqyKQq8yywLuTeKoqhL5OZSIPeJarlVigR1xVQxARzvFHdgjdlRxzDYlO9cfFftlLwBrpVPYIw6t rlxui4ZbO2VzSLYVl7YLrHzcPTezsEccWl253OXaW6ETFflVOnDbLJlj3GGP/s88Ku+UrTw2w7Bs I22XxM0sR3tSrtWVy409Ys+NPbKXz6oN5QcOq4/smGmZvyI3yxeIPXqxPbIXBazJJmxOeS90rGGP quoRa6yxWmbV/vJO2bEGe/TCesRFcfkR105Y2Qku7jVij15Mj5CtBrA2iUsYVsqjUoefC7BH7Kix WKtaGGKP/m89MuVzuFM2jePmTo06YqPtXPNhj7jmtoIDPbJToAVqyAWWVlcuN/ao0gfbQbnmsThb MRzV1Ko3s1oHY4+4FmFuSplkc6flR7asqb6YmgD2iB2O96jMdavFHutL2rIeVV9GzQF7xA7sETuw R+yoLWN29WXUHLBH7HCcR9xha+lYfRk1B+wROxwmzp4FteZbjqoBe8QOx6u0jKrye16I4LIC9ogd L5LW/wrYI3Zgj9iBPWIH9ogd2CN2YI/YgT1iB/aIHdgjdmCP2IE9Ygf2iB3YI3Zgj9jh8GdjErBx DHZ2JRAa7o9QNhzYHfYP8U8Svuo/StjXv5Owp7+vUOp/l2jkf5pw9f+euOuXTvzp9y6R7RdPrPd7 i1jgN4wY49eP6O5HEnX8ehL3fF8mTvj2JjS+g4gPfF8jFL4TieG+c4nOviuJRr7fEM6+B4kHPpeI Ah+B8KxPU+Fpn37C4z5xwFTY1sC+E8ICn7vCBz4+ImffEFEj38Gizr7jRcN9k0QKX5XoA1+1SOP7 reiE73eie767RXX8fhB199slGuO3TbTAL0O03u8jUbbfu6I//aaL7vq9LnL1f0XUyL+xSOovEPX0 vwhtyoG2rRcybeTy7E9beZhnaEYA43nG7XS67BdvK589HDSoF5qB4lA0mmLzmzkuOIeIOz5omWg/ 2guvUZKS2WGtS2YzfeAYvG906PHsQKf9aOzFcyhNddbwaoaApeSVb5wzPVkTGY5laK4jsJmxDqbX tPdnHtO5v7Q8Jh/z2Td+aA4P8j+5cxZ1hVemPOKu8fiWFsejRsbjGVjrNG93vOtjeF/Xp8lYxuPj 8H5SNpoTO8vwWNPSc8JoMJ8HwmI/85xS8/NIA63yWz6z1DIGIoHMl+ydfBH6B/r+VT+ExjgZP4uK Whq2cPTSMOazAbDPBfYNtvLPHJhjYbsfUsCZngpnOhH+Gw/bCL0M7Apkns06wrTdwETzfks9lm0x x7T50am2YppAtp93q/ZGaCp03k/gw9mm591+6T1duNYnwWaft/SxoufdWjzCtZwGW8+7XQGFREBh x+E1zfS82w98IiDOE4QrfNKEzDbb826ZY7g+79YlwPic2T5wwAin5/O822WBz+rtBZUMd3r23ox2 Jj8sn3frZQo65nm33UBggMD4mTmv+Xm3TL7a+7xbrMU2sBbbwFpsA2uxjRdMS3ZGWmrC6ISYQZMj usUMCq1OcoQWmsrPz8vNz83Kz87Mz84oyEovyEozp1RIVGaKKWUkGVNxRgKT0icXp8cYk8O0FFik /DKpoDSdKZPy8yyTw7Skl9VSgS67ip6HlnKiyisyaZnsKC3FlHWiqPKKzmSmMikjpTSVynGUlkxb WsrosuORyR3oxw7VUpCbWZCbUZDNpPzs9PwsY0ozpjOZaSZfTO6kGnwxKXKkluIKkn2PnvUgiHJH aMkFLTRTs8GXzArcSbfpTqkcx2opn+zbVNYdR2uh8rIKcrMq5U6pHBiSq6QlMyY0czSkVpAyIoIh pQ8KhJQGqbdXam/nlG6GFIqSDMmWU9buVFVL9uTeVGaaMRVkphZkMCk/IyU/PeVMehKkvLQEY8pN SwAtz9zJtnandCSCOavKWrLSrDsCVZoKio2poADkpISWlmDlTplRsepaBlFZ5ePYUg5llMNo6eZs 3Xcs3ClVBDN6lbTkJhi00MWZEcGpg1rlZmca5cDbiG5eWdCVzXJAS2o357JH23YHlhlV1pKdwWwk jc5P6A1yKHABwiimG2yDHFjNGLRQ0HegKxuPystIKZ9K5VRVS15SRLFBS17K5IK0BJOcM3n5SRHG 7fTUBKM1Bi1eJj/yMuEcUbnM/w1rrPQyWrKrpuVM0uhiOC+wkZqQlRRzJj3FKCF38iBQlxXTrdQa 0JI26JmWfAg6UJGdXlZLcXW0pIwuzmW05KclgZaC7EyQAHJSegcXwCBm0JWRmgRaINBh3CnVQuWa fDGIstKSUSUt+SkxxblZzEZGKmiB2IUNUJE6KLgAVr3ZmUZroAcxWiKCS7UUF5yBU2kxc0FYPUtV 05LKaAFTCpixLi0zKcZg1uS0Qa1AC+zJNZwpOHcwBmZYaDFEEA1yYGqHSwdjgrVC6TVE5bUUpE4u zssGLdBfIOUlROQbuk96RDemyoSIvKTRkKD7lNPCzFrgDmihzmRTedkFeYb5KzuTVUgFWqg0RkvF xzwbAGGGyBjdqlRLaWKmU8OUWJx/pvhMLhchFWpJsKvlmRyYsDLNWixhqYWjEHtackGOKeVmGVIm BDqMOzAMwqhsSGkGLTaqys3KMGqZzJxWrqhAizlNhr5TkBoDvRmCC2L9DNNTImBghgRTKCRYYJQv Ic+gBToudyEVaLHCmeysShVp1MKMzZUENy0WKTOvopRlTBnGVFkh3LQ8N2AttlG7tLDdE67K/aOB /gjFChEKDzDeR2LulQzxf1sYFhD33O4fdYG6B0Bh78DrctP9o64BA4TvBEwSdglYLmS2q3P/yHz/ lslTG+7fzrlsvC/KvJpRG+/fWus0b3O9f8v4bT4/QsR+/5ZA9u/fMvfTDomhbzghNNgboU1W92+Z z1Y7Ge7lVXj/dgxi7udHo9moHYpFChQD7wcg4z1a5tV8v5ZhewsNlvcQLe8bWvYvpo0127/Ocupf Kylj/2BezaiN/ctap3mba//i0qeEiP03AYcCjH3qzUAbfSrA2KfYfxMQh6YZzu2zfsX8JoDRZO5P zHZV+pf1/e8UOGgk7NDC2D3V7fnd/zbXmw0VTnZ79t4Mtvvf3fH9bwcAa7ENrMU2sBbbwFpsA9// rp4WfP+7clrw/W98/9uWFnz/uyIt+P43vv+N739XRgu+/43vf1dKC77/je9/V0ILvv+N739XRosV 8P3v/xhYi23ULi2O/jfRzP3vYl+EdjpBmfA62tl4r+SR70jnR767nbj8m2Fb9795qCya2TjW8v73 X1D3AtDQwA+hHCfjfdC/fHOcHvkucGrgl+PEbLPd/+ahiu9/m/Mw90XN9yqZbVv3Kk3an9u98HVz jPeYmVcznte9yqws7vcqrXWat7neq2T8Np8Hpn+azwOzXRvOA+OB4f7rG6bz4Gz0LafoLAqDfMwN u5B7xuOPm45n2s+chxLI64jzwJTHes/YSqd5m+t5ECH2e8ZOiP13CF0hZqPdELoMMXvcrew9Y+az vW7sv0PogxRoKoqHszsNvYOe/Q7B/O/HK7pPbK0nDXgddv4JdYLXaIB7WT3MZx3djcdUpCfWUNZ4 g54phn/ZPt50Jzux9N+2m3WZ72k3sHo1b1d0T7s647QKOs8GIAWD3maBsezVgo2CQuFGAZdxujp1 /+xmrLuJ57O6D7ltFEg8a77uX3yMdb/k/6xunc9GQQt/bnXbysOUYR5/mO3a8Juowx8Y45p5NYNt HjCPPwwc8ZsVLuOPtU7zNtfxx9J7xuba4H25Odg09rc8WfHYz7TZEWO/2XumvJqegxlrLf+uC9s8 wEfsvx1aJEboN3jt641QOK/suMt8FsJj/+1Qfxhrpxn+jggz7ioMf09kKCo7BwSjsmOt9d8Tsda2 EqiA+ofAjtdB214rbcxnX8MONzvamDmhF8wHzIwQh2IM6uJA3zTEzF1lf9fEsL2FRsvfPbH93RNz 37A1LgpQ+XGROYc/wiLZXWBct8cLjPm0xBSBlvASPPL1sjk2Wp7zitbu1mNOMxvHWq7d14OOWPjw PLxuFBjX7ueJWNCxUbAeXplttrW79e+2quqL+XrG7EuSwHw9w3gy+7l5ct50TgL8nnkS4BcPGjYK zvvGC5jt5+kJBTuWQYOK3Z95ctc9lX/X/fl58oe70RNPj2eeeHrEC+66bxT84R4vYLYr6wnzUzk/ U/1eFtsYNQcBkqFxaBIaCKPgBBgLKwsfxOcx/Yjpv7TV3F9xnc/6+0gYgRMMo3C44VemM2HFzswe iRzrbwD1l/1WhR2TIJh+2mVdv9kB81/CMs8M9tEYooj5TS5zbcS1fgIZY5+BEK5CmDYzdc6Aq6bK qmhpar9zJerPhfY79zNuL4TtRcDFwBTgu8AlwKXAZcD3gMuB7wNTgSuAHwBVPOPabhXwQ+BqYBpw DXAt8CPgOuDHwE+AnwLTgZ8B1wPVwM+BXwA3AL8EfgXMAG4EbuIZ5/NvgJuB3wK3ADOBW4Ea4Dbg duAO4HfAncAs4C7g98DdwB+Ae4A/8oxrhmzm3AO1wBzgPuB+4M/AA8CDwKfg4yF4PQw8AjwK/AV4 DPgrMA+oAx4HngCeBJ4Cngb+BjwD/B34B/BP4F/As8BzwL+B+UA98DzwAvAi8BLwMvAK8CqwAHgN eN2kx0xHY2RcQpwiqGXkkFZB4TOiZ8dNm+jhauwZImOMQKdgxuEGA/SS/lkIzeYZ1/sI/RN747Yg 8PodKGQxoh/coe8+KnLildxG7yXfuB1AwAo0CEb3ZBrl5OSgq1ev0ujXX39F3333HYU++eSTYJSM JtATJqAhQ4agDsoOHVC9evWQsGlhUeA5mnAt3YgKLCyi7xTtLizqeu1u2L/3nMR1xRf+fXD7xu26 l0qeuBcW8VAkrOvoIFjazYiLnnIjjO9dcM9rqJDPR1r0GrQndi5KgL31C+75T3UZophGKxLfiR8f NELx7YzEhYKXCu55DBWgp4n8Ee9MHaeYgjpFFhaR1+4eJ+/QhOivG7eDrt19PaweDFd1eW1RKKwy o+oZFIkKixpcu/s5o6iwiNEzqOAe0b2wKNwgKZR37W7yndt9xa+J79w+/6gt78btt28U3PMsLBpW WMRHke9Cs7rqH6fyUnlX7/O2roIhVNCaENYjnDYXFjkXFrkUFrkWFrnBG2iiR2HR6ksl0vU5/Cge j7eHp+fx5WgcDz118eGJA57wxQGCcdBL3r23m7hx+73HgpC5flvl/S/MTSF3CcKu3nfV8vgdfb1W 8cMDfPzbeV1uWGeij2BEgI8vMY6/2dtPOD9AND/AaX6As88X3kv2QhFRj4kdrtcCw1waLajXos6n fl7+m32W5hD8G7f/fCwY7FtYNEA6unkn37zRTc+TQbOD7hRRvYZdDpzoO7r5cl/emjb8NW1WCda0 CJa+5yeaAbucvk5qvt9/gb/L/ADXtt6f899xPv9Y4Cb+QeDit8B/Xl20wL9BqwX+myV767uSDdq7 +ES5zQ9wnx/gMT/Ac35AlHh+gNf8AO/5AT7zAzJ823p/dRgU/tOOfFIymd/J5yxMPtHo6eZXevPC +/LD+wrC+xKtw/sKw/uKwvs6hYelkwO2vnKKvNJ9bDuvV3p7h/eN8gnv6xve1y+8rz+8CQjvWye8 b2B4WEo954HewbJOPl8eh+L7PRYkd+j16us9e7j4Hu7l66sYyXdVjBQc7rXpNBHjPyBy4utx3/8B +YofC4KCRgSkHpGPCHB2PisvkT+oExTmHdBwrKbJqjrvN5zoQ4RMaShc1UA0znll5+ljJ/pE3npt ua9LyBTXkCluq+pczz4HpWwPfr2IXLbUd5lvZ7LVGJc66ibBy2+8xPecJ/BM7D+h3TzhoHbzRO3m ObWbvaT/T+cJ1xu3hz8m+A2FlMK9/gavxsOG8KL7nkyM76W9DIXxTgbLCVVH/1dkw2TjZImysCWy rWQO+bXsMvmgS841Yv+N2+d+RA9m/sgb4ZdDHi1Ycbzkgz9c199alhDFD1l1pd1akWqY2k/VXNVV hQapxqimqZJVHy708l+v2pYSMdHHq+Ha5b7e7lt8Hi4ZtNzXt80WPzfY4T94S6cA9y113LcEBknX LPf9pm7qJhefevMD6rf1jkveTfzI+/d2GjFP+J3L+TroF5R2qWTvXeV4rxbIIyTkw6iDg30exo0g x5PE7I4D5hbN9FvR/m4T/12zVkc5J65xSVzjmrjGLXFNK/fElaIvQz4MmLSX2K/xO625VLLmUsnh p5tGS+vkfhrcofX+u9pdAJ5X+/Bl0k+kW6RPg4a3iW99QED6Dozd9f3omF3fL+vo5Z/yU+DhXpNz iMPav7SXSpahSyVBr0+MauvdjPejMFQ8SJ887OTPvLf+Pv/ohH+KsC2vZdgqp4ne8R7Xj/zoQrm0 eyz28pQE+b23NrRu21ckT18fE5jS4/tf6zb5q+HlBj2+79TqU9Rbmk5dfDhDmiIdrZui2yNNnk4u bp+z/SfNNd017ZUHIxChW0i31HU/1M6r/5FlXZrx0IGsbvpb9Jt6hb59YIrwZx6v6SqnGzM9fvuZ F+Dyo1CnRxf0d/VCqg7ViupBDaHkFJpOtT595uDfdB/JQB/S/426SChZ9udZ3rHzn/62JuqU25tu yQrxQvFq8dJ1nYJbCjs1iVK0b3Boun/QFtJPmty8+RSpjvr4ZPPm1OXoNnx+dBtBdIu2J9JOdNKH SZ7QSCyer/c6GTh81/b00+oToh9Q4PCP3vn5z7QTgb1yJCcllyVNM8ReJ/t7Nv32T5SU8afrmDea 1nnEG0dM71znltgZSXwkKYXTPaVBTd6nL3eIfEX62z/8T44JNh8jfprO+2rnH6u8o0K21d1ztyV/ dhqPd2VGr1DvE7LgqJCeUQUyr3HqEzOP9EpO4/2g6Uu/RvOpb19++s33v/lsv/ZuRtKJ9kTCTo+N R/IGLN73bZD84fcnaWdY9owZExlFjxg2IiQkdEw0PTc0pG3izIR4akaHkC4wtBKIRuloipPgdx5N 8OAy6oarIsk5VgILtsd0I94jHoKLu67tOgXRc4JCQxsHDR4vTkwcP+Mc/brYC012/+ktGPE7oTmr UZAWDUN90OBVaPy1eeKd4kslgz9GkShUgLrniG+Kn4q9JI0lHSXJ/SSjJJMk8yQfSD6X7JCg+QOo onNQxYFfvG5O4+V6d3S7XPzABS1Fl93niJ9I3na6cTt2Q2GRkJB2lOofw3TqxWsxJnh0sGfbmdKl 0o+l30pRtlQnvSC9KxWSdchWZA8SDSHlJPQqcg25ifyB/IVEf5MUyZP5yprJusgGyt6QoakypWyV 7EtZluyQ7E9ZoQw9kYnlwfIO8r7y1+Rx8iQ5WiFXy7fLf5b/Ji+QP5S7KVBDhVTRSzFCMV4xW+HT apq0ZaL4B0Vgi/aFRWtzFFceoI68es14Sl9lM2UX5da4rYqIt50EI5RXHkTMiB/sizwUVNEN13nK TKVWeUKJhinvT38yGGYy5I1aI5qHYGGREK+gYX5lTtcI2nS65tKm0zWDhtOFQjuEtqND23dpF9qt WztJlw7dBw8791TYsbCoJSwFxAHOwqDCIj/YFEX5G+bghTCdXrs75997t++cg5VvkVYsHtZRDAbD dNyv4B6/sKihcToez3tTcOd22DtiN8Fyj/OPopnpeLRz26C3+JEwETt/IdaINsPYWVj04aWSxYVF AhpF9XL9/cv4l+nNzcJ6+W3tf0E/d6d/GKwLcsiAQH/twPoE2TqA4pGSkuGSGMnFhzxagIhNsMy9 CWPwKuQb5n71fnrDInJrwy8DecJC/iGdwAt5b+NTPB7RNuAtHjEzGo09wLt+51vh0ws+goIA4W0f 4cCZ7kGP6o6/Bo1rcNDp8yVeM+svD3wkhBkc+baE5UD/63cG3CTCgxc2mcxvHh4sQLOD54npxV5F d9GuVlfcwptRvN3CTtL+0kslp9/zvnr/4dp6+l1+nzsLw3L8qF3+/Tt36tQx5GS9pBzBbcl5Uvi0 jvSTd0PrvcXjJ5Z4zw4OUr555eOS3dSu9E4jBImNgjd3zU3vhDJevr8xvRNP26V9wG8hh3p0vbgk 5IJvaDMescWpiUc/4ZSbBCyEQvhv8aN+4f3Cu/hQ2kvaq8ClZTTPPfhRsLCw6L1LJe7Bb0HjeeLl ywPHrpU0vS5MaI+KGya05zVse8Fnyy3+tF6u6IIP70I/Hu/9TOdPe63r37/zt7K2vj/LEK/Jo165 zZzkvTsGydvLF76GHvV61Gz0G18GTJdny9a8Oa1XxI3b7bfKrwvFb/FnULwb8UFP64/jBcfA2uiK Z6Zz3YOCq/f7KKbS8dHT3gkKV8QoZ04dPw0uqmLCW96UPgWfYCpqogzf6CqHySgkat+4g4E+zXjz oofKB6xow/9Mej8WRfME86+45TaLBm/nRc8ORoeJC94XH86LnhcXNz6aV7/bw04UIqIHjdJvSKHv o7CwKfy87540fzn09/WLEv9ptSDLa8SDQyHjlVlerdcnnM+YeT6mc/JUUkk+XkV+2a35cGfvtoVP vt9FBtVd3pDY+Cju8HEyaFbj8ROLpwceuUEGzYyZ5XbqNX7vdhd9onmFvtG8gS4TeHGL3CfwFyzh xy76/IqPd5PEJVd8Fi3iLVgiOLhueeBenkLtt34BH/XiH5m+fuammRQ6PoOfJHf+MfTym31jZsxZ /DMfNe07Tsjrs1ow6OMLvolLbvS54hMwZxEvPdP5+q8p22b9rXLekv7NupP1wrzUjdW/vfHraOeL s44OPZ4oqa+pH3QADfwmRPHRokUB4QuWCJfKRiUF5wu8mAlG/fkWctt6zbaN1+4G+p0nD6DzzZm1 4XUwzGNRy+mxxPTo63cWzFXMG9/v88vFBxDETORxnff5L37fXu9C/pNX+F5P6WCX4UN7BlzwmvTu 00+d7yvQSu0G7U5trvYP7Q3tY62nDgXp2uvCdSN1b+vm6lJ163XJ23T7dad1V3UlOld9A/2T3neK Xo7fmbB9GnREp2MEMWdhfLCbMuXKF92i60df8I7yLizyKSyCRaxf4b0Nx6fX6RHeN1odWn9abMOI wqJGdXce4AUXFjXOiI2CK5BmhUXNC4tavJIY1aqwqHVh0UuFRW0Ki75pW1gUu1AaEh76JLIDHNRx 4fTCos6FRV0Ki6K6FhZ1KyyCy4AeMO707PhHc6orNYgaQ02jkin0IfUVtYs6TP1F3aSeUl701oYP I5AipCjxZ9S3B5rWsWgGeideMoKOuPIgYGoYHec6PDJnhNuCvxNu1c8bGngweuWt5XV4qE9I0VQU Txq/t3mT1tGLeg3yKQqZF73EhS/2Ew/56ZpIhqag8beOh3Q/MTa6Y1EcmtUR9RMTo8Sx2+sVhTSb +YhOFXeb+VL49FbigR2Lxu8s6sELnRou/lP8+Uk+r+HDv8XIS7Lwi5IoRZaEGiAZLZkiQQskKyUb JDsluZI/JDckjyUIJnhpe2m4dKT0belcaaoUrZduk+6XnpZelZZIYR1PLuk8lhY0H/vSbeWT/ieG qXvPmJnwdr+TMvKeoNuOr4mZ5FLyY3JS7KKFaBd5mPyLnDyuThJNesuayFAnWX/Z67LJsvkylewL 2XeysIOy32XXZY9kHvI097CvUf/eQxp/TQgu/HIiet7xhzFt+HOiEboZjwToQLexrw+frZ0hT5Gv laOv5Xvkx+T58iI5X+GnaK5AXRWDFGMU0xT3XOYq1iu2KdB+xWnFVUWJwlXZQNlOGaZEEcpY5Szl MuUnyvcUPymPK9FF5T2lSBWoaq3qqRqqGqtCM1QpqrWqr1V7VN8o81VFKsRX+6mbq7uqB6nHqKep k9XoQ/VX6l3qw+q/1EdUT9VeGtRY01HTTzNKM0kzT/OB5nMN2qE5oDmjuab5V3Nd3UgbokW9tZHa Cdo52uXadO1WbY4WndRe1j7QOuvq6dro6muH6dA4XaJuiW6dbrNury5Pd16H7ugIfYC+pb67frC+ hy5ejxbp0/Qb9bv1R/Xn9Lf0iEI+VFOqMzWAGk1Nod7Qr6TQBmonlUv9Qd2gHlOedBCN2tPh9Ej6 bXounUonUdtotJ8+TV+lS2hXcQNxO3GYGEWIY8WzxMvEn4i3iD+lfxKji+J7YpEkUNJa0lMyVDJW gmZIUiRrJV9L9kiOSfIlv4oRX+onbS7tKh0kHSOdJk2Wog+lX0l3SQ9L/5IyQ+4/EtSY7Ej2I0eR k8h55Afk5yTaQR4gz5DXyH9Jd1kjmQeJessiZRNkc2TLZemyrbIcGTopuyx7IHOW15O3kb8ibwtd X54oXyJfJ98s3yvPk5+XoztyQhGgaKnorhiseEsRr0ByeZpio2K34qjinOKWAimRj7KpsrNygHK0 copygXKlEikVO5W5yj+UN5SPlZ6qIBVqrwpXjVS9rZqrSlXB9Y4KqZWnVVdVJSpXdQN1O3WYGkWo Y9Wz1MvUn6i3qH9SH1cjreqeWqQJ1LTW9NQM1YzVoBmaFM1azdeaPZpjmnxNkaZEr+ZrG2ZJEEK8 xyHbOu4IYr7/jEQNtSO047UeWUrtBz9O1W7Xop+1v2kLtA+1brqGOqnOTYtG6MbrZuve083RZurS tSd06JLuvs5JX1f/kv5lvUz/ii45Qf+u/iP9N/r1RzbkHNan8FLOnfjl1C7dpzp3SqsLoVBvKpKa QM2hJurTqa1UDoVOUvv0p/RX9H9SbWgfehiNxtGJ9BJ6Hb2U2kvn0edp9AfVcQfvrqsLRb0k7vnv IDFqL36FXiROE28U7xZvos+J0S0xkhwV88S+4n6SWPFAMZonmSp+NjT8KTYMDf4waERKnxTES7s0 QPOkX0p3tPpE0uxRiPS+1IlEDyQvkS+TMvIV6XRynHQNiZZImXXsXmme9DrpL2shQwHkq7I3ZQrZ W+Qq2Q4yS4aM69gnspUNlzScI+sjR1HycNk78vflc2WpsvWyE3J0SX5f7qSoq3CVN5C3k8co0ExF rPxjxbeKbEWzB00VGhL9q3BXNlKGKHsrI5UTlDCApirXK1co9itPK39W/KYoUKB6qjaqV1TDVONU iaolqtlKtEX1kypTeVF1TyVSZ8q1cvS+6gNViTJe3W5akjpD/b0abVD9rT6m4ml8Nc00fur+GtRV PVkzRq3SzNYs0RzU/K5Bh9VPNDfVwdoO2r7adpqXFWiU5n3tRc0i7RUNpfxJ27kE3dDe1bTU1dEO 1r2li9fJtWi69qp2l26T9i+dl+6EjtIiQu+rG6DvopuiX6BfqVfqUFN9j/p79Jf1HRRPyA1wdYiC KKk+lppFjdd/Qm2hfqJQpl6rP6En6Pv6UKon3ZJGL1My6mV6Lf0utYcm6bUH0BW6H/2h7g6c6gCa omPEaKZ4qTiK/lacLd5CXxAfp9FFeoVYJA4UvyrpKc4QL5agNZJNkrXimZK/JcfEPOljZ/RIHCLt Lf2dunkjXLJYOp9G30h/lP4qHdTuTymfvC1pTiJ/aQvpGPJV6ZtShXQDuVqKmOH+JhlH5pCNZd5k PxmSkuGKebKDkmnkBNk68rQMfSj7V+YubyQPkfeWh8omyNEc+URZuvx92WcyjWyf7J4cXZHVU7RR vKIYphinSFR4K9A6xUL5X1F/yVfLvaa9Q3qqUH2lSNFLOUI5Xjlb+Z5yjgIxlyZbFTmKk4rLigcK ZyXqoWqjjFYlqN5VfaT6RvWjCm1W6lW+qlNjOqbeUWbK76vQeflE9XjJCrVavV39s/o3NTqgeqh2 0/yrclc3Uoeo/TUoUr1UM0e9XK3VnNBc0tzXDLusrqt9Seu7XZL5Of/rDNIY7T9KBmnQ6X/6aXQj fTXJez+CSeCaFv2rddc10oXo3LWRugm6KC2sdd7RbtFd1Gq0+7SntAd0qJ6+jf4V/TA9qUvUL9HP 1KGFupW6M7d26Q7LedQdXYQKdaNepd6kFNRCKl6fQaXp0bMpQEAxMwDaQU2m59Mq+gt6JXWQ3kmh 63Rj2jgF/EszMwAaSb8v/kysEe8Tb6OviP+iUT+Xv3VR7p0l/UQBdb3EEjGaKI4XZ0i+lxyRnJX8 I4FBwFvKXOfPlHSXWKwPAqW/S69LH0k9SFgfhJLNpcx3OgLxYum3eQNJ9Jn0OLlP+oB0ltWTtZHV J9EwGUnOkKXI1ORm2cfktyTKJotkF8gAeUt5d/lgeQ8ZipfLZX/LekjXy9fJxkvbvI8uyt0VT2Tt FeGKkYq3FXMVKE6+XrECpnnzlSx6KG+jZK5kzdM4YuZxne688o6yjWq/7B6NGqj+Vr6lkinN0zhi 5vFflTEqH7VA5a9qoUKj1U9UC9QKlXkaR4Wqm+p/VNM/WiblnVmqTpOibuokzRS1WkPomQC/rkGP NB7ax+pQbZCmvWaNZqQGva35VJuq0Wq3aa5oi7UuOlRf11bXQDtcV6JK0L2r+0iH3tF+oN2i3ag7 rhXq6+hb6VEP/RB9T910/Vjdan2G/ns9OqLfozumW6kv0vH1XaiBFHqDmkqN0a+ivqSe6g9Sv8Op pB5RfkqCVqmZb0viaJREr6DVdG/6Z3oHdYUuppGLuL7YnT5Hh9A76Nni98ToU3GmWCveSl8SH6N3 yOpIUCtJD8kIsbckqt0dMTORI/NMzkzkAskECRogHS2dIn1TopAslDATOTLP5MaJvLkUtZHOI5dI Z5OD0/eRp8grJComXWT1ZW1lDcjhshhZBImWyj4nl5GfkDrZT+RdmVCO6shbyXvIh8h7yqbLF8vX yNF3E8YP17y2Wb5rzEHZJZlAgfwVLRTdFK8q3lQoFG/JVytQhuJ7xUb5bvkpxTn5LXkTJeqk7K98 XTlZOV+pUi5QfKdEB5W/K3MVmTMGfvxoTt/P6lGom+pV1ZsqhWqharUqQ/W9Cm1UnlX9o6JVt5RI 5aNqqkIC9WT1fLVK/YX6O/VB9e9qlKt6pPbQSDSe6iB1e3W4Go1Uv6/5TKPR7NOc0lzRFGs+v6qu r22rJbUB2yRzPxe4zswIem17ZKimjyZKM1GD3tEwwWs868xJRy5aUhejm6lrq7ml+EhXrEBbtBd0 d3XGs86cdCTXD9Ut1q/Rp+i+1X0PJx0d0/EoX6oZxZx15qQjJTVN/yWVRR2i5ssKqU0ytH/oUcpy +YbW0yso4/LtN4pZviE3uiFtuXxDP4kzaePy7bi4pRgWZh0bBUqmSxZLmDH9B8kvEvS3hJLki32l zaRdpM0lXSVokGSMZBVM3lnSQ9I/pYVS9EQqJp9KOpB9ydfIQNJjAZKIF0q3kfvJ0+RVsoR0laEG snayhmSELFY2S7ZMNptE75GfksdlF2X3ZCJ5oLy1HPWUD5W/LDOu0ZklOrJcozNLdMSs0ZMVCvlX il2Kw4q/FOim4qnCS9lY2VHZTzlKOUmJ5ik/UH6uVCkOKM8oryn/VSJ3VSNViKq3KlI1QTVHtVyF 0lVbVTkqjfKy6oHKWV1PjdqoX1EPU49TJ6qXqNepN6vRXnWe+rxapyI0AZqWmu4aNFjzliZes0iT ptmo2a05qkHnNLc0SOuj5Wk6awdoR2vRFO0CreUFKWKuSJkL0mCt8YIUMVeklhekqIG+nT5MH6Hv pZulX6b/RI+26H/SH9df1N/Ti6hAqjWFelJDqbHUDCpav5b6mtpDoWNUPlVE8Wk/ujndlR5EozH0 NDqZ/pBeSO2iD9MwQN+kn9Je4sbijuJ+4lHiSbAGE38g/ly8Q3xA/B19TfyvGLlLGklCJL0lkRDA cyTLJShdslVi/Pr2lNhZWo+J2Vekw6TjpInSJdJ10s1SxKy+zkvvSAnyrqQl2Z1Eg8m3yHhyEZlG biR3k0dJdI68BZedPrKmMl9ygGw0XGXKFshWyjbIdspyZX/IbsjQY5mnnPmmJlw+Ut5XNleOUuXr 5dvk++Wn5VflJXJXBWqgaKcIU0QoYhWzFOMjViiQWrFdYVx9PVS4KRsqS6RKZtpmfvEiEzPXsSOV ycoPlWvj+7siftzotcojSsHReJ7fURjSCVWAql7yFuXLKiRTmafy1jCV/6pCetVtlUDtr/5a1U39 qvpNNVKoF6pXG9ZsR9Rn1f+oEa321jTRdNL017yumayZr0EqzRea7zTGIZ0Z0YdJtKHaPtoo7ZSR 2/gjt3USjdwWErcjKnDktqTtYSN/iN8ev33s7EnfnZtOjNx2Xzhy23sn5glQFGrBEy9P3zlye1jO d79tP318V0xiQNyOmVO2jQ3qPDJL+LRoB2/mufh/f3j4ftzOEz8YviWa2KqIhBFjdxFJ6IoCDp2b 3lTXOYf52kg8XllE9tKN0G041GAf+nb7Ce2xnFRdiXabrsE+zfago7pL2lu60TrZ8QFf7X77BMVD keFf7xSglsmt0t//7dz0j/9e82PuQfX2t13X7W4QtyMyaXvuwY9++DVr88jtn47c/bDxX/ouB8L6 PuxykGp+q8GV+ueHvfnOtzmCnieJ3n+e/+agsKP8yJQL9eK28xP7nHj3BB1k+n5w4+IrgsSx+1Zd y118Bd1I/Qfe8zYWnJzL05xN33rymxyvP6fzI89Nv3OSuHVRcO+q8O3tyert8dpF2jTtRu1u7dHd Y5Mb6zvq++nlv966eHZG1G1Xf3TrIu/wA14D3cL77a6iD+mvYMS8Rre5wdzt3a0Le1n83oNR4q9o F/3I79qvEd8YN1Mct2NG39uuCN1/yvfhOfvy0FPPO5dK6v5xamRWH8XUeNrwhV90YjQ5LjphPIr7 M/GU3u+KfuTutJG7m0y4N/L7Kdv0hq/8vvY5SAbFzoj/bfjtKdsTz/FT9Gv1YV/rwx/8/PeU7Xt4 Y/ycR+5dOWV70MFe9/nXg/sWFYrCZqOwsF93JDuH3SJcbj+5v7j5hRUtjKuwxwup1edXPprW6l6L 6z98Rt1oeuR7KmhWvJ5Ydcbt8AkqqJ+20zHPD4IuKlDT49TF4bdLLg3s5DqgsytxtE97wcDOwhGd p3R2FfKWHN3dxjuu6MGfve7v8PG8H9a0RdCdDtK+0tfpg9tVR3fztl1zXdPyXWlTGUzlLZu6TtkX fOsCn5/FH8fbL+QJfmmrkSb1O9pkee+3T0TdRvskpyRXJMUSF+nRM+em9wntPcbP/VfljNygZiPa esY2pUW7mj1M5t0jlWF7p25a8dHeuUHKmb6xAfVF558U//th5PEnxcMmf3fzLapntyTt+OnLBG1v JvfpO2zsTA+vscRl7Tg+cprp8etwt2knRwqSvT9ETxvHbX9ddHbUL6OSfW4Nr7NNEHli+C8RgjPD Rb4xvPMRY89GNI3NkO2Tzh4o4EfdblNyPXGC8LOjuxvFtpqK/r0lyJkb5D7l1Ts8WlBChs05O7R+ gfQH54SgPf/8fPM85dw48/vmZJCkWPd3xDUZGbrkH/dd+qXtfQ/suyMN8pHE3xjy9yCyc+SdO6fC EyckTBT16QArhAW8xAlB82YPmL944hLlsPh3L5Wo/GjmDjlzg/yMxgfxbokEw4oESMAb1rfI837n 14Lu9JW/9kabvvfvkFG8Qz+2vibmEQv5I2SBv458d/VbCtUUwVDke++P28yXUgWLpmy/v2LgSyjh 4JTtv0AnurBwjN+tCyN/+Hz47Zi6965M2c6/3im2VRDNCwv7g0ifOISX0+x8waLLOfm5qrSPh+l/ fve8unnyN9Khil+ly6+31N5rUff9Sbt9lEGk4fvguGadtdfIuOzkqUNfc+v85zWl68cTv17ucWpC WHA76mIAdKCA2y6pMwJuR7gv5PnfSdzAH7lBMGbjCeGEje9tSNzAi98oWHfx6O69r6k+IFau5yM+ +nFVyEH9oYg+KX0meqQNmsbr3cZ3xqQvWzT+iWjjw+v03UJBzx9uXZqw8cTd9zZM2sATpl8qua79 asNanWb93o9H8T9vHcbX+mn3uo2SBYn3C52ImRIP3bmVe9A4RfzGLU8ux28ULu3s3H/C3fZdGhya vipnC7lStyF3i7rDdwfOk3vQ+SdhFK8Xj7h14dywPWdG+Y3yVrfftyBakirsmqPusgculWjjF8Lb 6/1zQd0TvXKo6Cnt4tX16YMzt86Pj3r6JEzdqUBl/6uuBdRKKrnsV11Pej8Je3nc6fEQDgn7nH7p SEz6wsMLXRjmtuj3fUHefO+dJ2//MwyJxyJfdYgfXx3irw4R7vOLlnu38b61qm/9KL+GH0W7nd7D C1aHfH1nuMq7T2gzdUhzdUiL1iFprVamvKQOaaMOOTh888gvpOqQEHVI6GWiQ4Q6pGPCKnVIZ3VI F3VIVFd1SDd1SHd1SA91CK9n85JAcWtxT/FQ8VjxDDFKEa8Vfy3eIz4mzhcXifmSreK6EUjR2Cux RRDzvXBzrxnonVZ/95FEqLsG5IXdfs+1XlqOusOTfupu5w3fDP9wcMS2V5fyUJ/GXlNpFG/8ZthZ JtknUXf3aZQzdJXYXyeJ5EvVoUMGN5fdix9/vkgfgoai6OZecSh4VvMg5ts6dWi9RjkPZ476R7JA +tKa8On1pQObe703fqNX2+CpnaW/SD+8w+fxxHV1cFmfHxp6+0anVveRv/gVchg5jkwkl5DryM0k 2kvmkefJOyQhC5C1lP2Pr/cOixr9wv7PzAAiloAillUDVrAFXBu2jB3d1QCC2EOxt1DEspYM2NcS UOxosODaMKBi14CIKOsaUEGxELAhrhpsKCrknYfIwvf9vdfvz3UvvNRnnuecc5/PfU9vCn6nJlPB VCQVQx2izlGZlOdj6h0FNDZB1zfL8Iv/V8wrpIX7e0+kDrvP6PVhAP3J/eSuNtNpQNLOmfl1N56m r9PYjtq74T1tYGqOBBDBbGHQSHCTecS8ZVSmd+1Oxl2OA0e22tXm8/vMwoiA2o5Bc05aRwDgb4IB rvmW/BYyU4aqyfA4iwTepyz8r8D70gXmcFHcfu40d517wL3mfnBQNQqM4Wfwf/AbeFjCJfJX+Xv8 S/4rbyM0F6CLYBS8hSnCAmGtsEs4LsBuPkt4KnwSrMQmYgexrwijRH8xVFwpbhMPixfEWyJcFEpE vWQvtZN6SSOkiRLMk0zSZumgZJ7upYfSGwkeibayVijHyrPkpTJskuPkk/I1OVd+JX+T6yrQUqkn V4l8fyqxCiCV747yXClVrNVmaicV+qudlSqR76h6SQWk8n1Ay1TzDNAb+x2DydjI/wS+TOynwNcA b4P3qJ7tawp8/2/tfyOeWEP7hy6kkfQmp5ALKoe+4yRcIbNIkfhEWlFNqA5UXwpGUf4Umv63UYep C9QtCvKpEkom7el2dC96BD2Rhnm0id5MH6TP0Dfoh/QbGipoW0alujFDmbHMLGYpA5uYOOYkc43J ZV4xSNsDJO7hDBL3FrF/srEsnGBT2Dvsc7aUteaQtgeauOfFruZ2cEe5Sxzc5gq4D5wF78A78735 33mYzAfzkXwId4g/x2fyj3l4x4PQQGgj9BA8hPHCHAGWC1HCfiGavy48EF4LPwSoLzqKXcXB4hhx hviHuEGEvWKieFVMEl6KX0UbqbkEXSSj5C1NkRZIa6Vd0nEp5YqUJT2VssVPEpaDQ5G5222V1e6O I0AtA/hg8hDZT1bTQ2QwPaLlA3KynCHnyf/K5TJgipNSLg1R/JSZyhJllgy8sklOU3KUIqVMQTMf EOoA1VWZqi5U16m71ajiujEFF9SVK2+/v/zxsLJRgQo1SakeHcapUD06pKv31b8xaIlb1BgdFmOg jQ6ZWLs7Hw0/mhZi0ILo0qwv0YZwxbXxAY4SO3FtfLhEfMQtCehFjiH6EPNImtDGB8gkH5N/E2h8 qGWeK4ZSoNQKpFz6zqO2UgeHbyAb/wKtqCKqjHpFtqAJegDtSkEQ7UOtoRdRx+jLdOUa4FGlwGtN 92MoJoDxpGEFc5D+iznP/M08YRQmsh8s6jeLQbJADwbJAnMYWM5EMWlsDlvElrF1uB8M1GcdWV9u OjeGXc/t4QQOGjs4cPvoN1wFh7ad3XgYyo/lP9DL+Sie5U7z13lI5jK4PK6u0FJwFQYKPgJMExYJ M/m9QqLA8/eElwJ8FXg2iV0mmITXfKDouAnmitvFI2KMIIlXhI+ipQSNJSvRXeogTpJGiRHSTAkW SWelm9IFUZHyxYZyWxl6yo4SwY2Qlsn3pPnyfQkK+UTZuclj+YXUTKkt91fAUwlUvOQg+YF8WNkp 31JAr6QphXKpYqn0Vl2Uyea3X41UQxQHtXOf42qu2pYDhY5Rv6gNMCd1DDYD81NhA7YXS8R4NUlN U0uxIhVaY13wZjiBDcAIfC2+EIPjuBu+9vl98/y/UnmOtSDAGi/EfYnpxGJiGL6HEMzXHb9LXMXv 4SzxFbch+pmvO7Gd0DZ6a4nppGS+7sRH8p3zW6IV1Y26ieXbQA8ynGLw3VQCJVJ9R/1NwSfqGdmE rkU1pZAaQFEQQMXQqyhtYzeBPkmDPWOgezFOdA9uHnOW9KdhLPMnfZ1ZyVSpAdCNbc0gOWAcs4ld xnAM7GNOMS/Z+0yVGgBIDjBwf3Jh7K3wW+wqFvSbZtMg1OO/sr/yQ3g/HmbyS/hZHBJ447iT3DUO crlX3Deus9CSHy1MFRYKsE7YLSQIsXy2YCmkL213HJ7zPFsk3GHHiX4kK0aLcEBMFjPEM8K/Yrn4 RqgQwFZsJdaShoqLpVniUjFJMqVJOVKRlCvWkVvIrbLxb3yKfld6pTas80kg+0rX6/WSUhdYSqGP YZ25CDyU38gVsq3SSqmQYKgyVhkmR8qz5b3KPXmfDKfkdPmMUldtqbqqA1U3Baapi9TpSpgSqdyo f1i5wMJH9bkyWOiI9cMoLAALwyBQ3Y5VlYDKPc/nyhJwENNUvS04RGJn8UPYI9wef4yhCgBvfpaA ZQRH7CNOEbAfv0/cwnu5SMqwTs5kr3Zg7a4n7IhxRCCxnTxCXiThH1ImbxMGCmkI08lOJPQmqzQE G+om9YiCt5RKvSNb000oBCJ9xiGc2vOuD81RV+lT1Cv6Gw11mZZMPXog40YjhSCahlhmPb2HFuin zF3amm3GQie2P9uZCWS9GInpTEWx8CfjR7U8do+tYBWmDdeDAw9uPDeHm8BGcay5zB9gIZnNYPPY f9mWPMY5cWN40Oo4KuOpyh3+OQ8thdPMS7y+IPGewgB+gQBaHUdlXOR9BQsRPvO1hKbC76IiBIsB wjYRtDr+RMgXZSHo7B8UfHy5WFxNdRTnSpPFaKlUAXTBH0lvJVV6J7aWG0jQRlojeZjn9Y3ycilJ 3i/BfblY/i7XU3ClvjxIeS3AVGWhsk6ZLZvkvfIO5aoMX5Ta6i9qZ5VUuyhBqrcCq9Tt6hH1oop2 PJHqUwU+KS5YH2wkRmOj1BXYVgxK1LPYTewR9haz4kuxCBF64sPxCfhcnMWj8W54Mg4Hsft4Mf4d r0dUYLfxVuZfwGcSS4iNBE8kEXF4DgFX8INMbfIXsjM5hDCQw0bBc0Kr5FWFHKyoz8RYsjeFOjyK DCAhjNQqeVUhB1TJm1AtqXnmJ3wm3f8SnKLT6ft0Mf2drsfgTH0aBjG+zGB6MbOZ/oPeQKcykEi/ YL4wtdlf2M4syXZhIIgNZ9ew8av9QvYtjGUPL4WzTA7zma3FNeU6cv04ioMAzpPVNjg72KNsurmi swVsI749r8k+DA8RfDCnyT7nOH5zn/PwdkfPy3WxjkI/gRIChDABVgnbhSPCDv4fQRbeCwU8fOCR KvhZmCQyYoS4RYR48ax4UzwnvBVV0U4CERqIbUSkCi6TOGmfdEqCdOm+VCw9EOvJuOwm187yxeek GKanOw7P9mktQXdpmDROmi2hu4sOHbRT/y65Kb7KdAWXoIBbpxRze+W7ygsFHTpop+6lGpVwdY26 QIE9ypHKU/+oWmKNMRcMtFMPwfzVrdhf2HkMGOYJtpM5Pe8Spt1dQP0bWs2wmNa+QQb2A69PlGMY rrVvgPo3hG7wuNa+wVWiGRFKtetvQwaR6E0HtJu5bH7UC8kswpJqTIEL1YTsQPYlR5ErzMX7LwrO U39TTyiF0tElZFu6Jw3DaRta3WRHhFH7adTAwwP6Nf2Drs84MhiNljeAtjeIQ9xIawwGVEMY2vIG avboqEWH6h49lAtgt3GAtje3OARhaAwGVEMYm/kI7gwPaHvzhkcQhsZgQDWEcVLYx+cKgCCMuiKC MDQGA6ohjDtiqlAqggZh9Jc8JY3BgGoIAzEYH0XQIIzJcrCsDaZQPZmiwbShDNWTqTaYQjWEgUT8 XxXQVPy9aqKqifhQDWEgEX+0CpqKfwXLwjQRH6ohDMRgmB/5KgijBNcYDKiGMDQGA+JxbXVjS2qr G6iGMDQGA9IJDcJwpTQGA6ohDI3BgBekBmF40hqDAdUQhsZggCVdBWFEMojBgGoIQ2MwwIPtyWgQ xn4WMRhQDWFoDAbM4PxCWS6aO8BpzVfKv1w5h/FOPCrbAKwOKhknD75XKL+SXxtlA/oJf6zlbS7y hktR9pfMT3opby1A3bi9PCEMELRSvk6A5uZSXrWrrSXuEqCj2E+kxAAxTFxV2bLBRfEfURbfiwap kdRecpfgN2mSxEgR0hYpvrJlM2lPup3cWu4uD5PnHPXI0ntkWXlkuU6446tr4pE1N9sjLzA7MNt7 +8S7R29HW3hkWXpkrUuLNZDgizDITfc8svcZT97NyL6a6xvjMOFO+JnJWf6OHjmWFU/v6NaF3456 k7d+wr20vLmR7K8lpPm5eFBClsrWRV9vRzsozgWVPCG0L/hVGaLEFNUv3JOdJl8pgOXKa3m/Ur9w X/YlJUcuUGJ+VwaUeHhk1Zpwb/x73ULwGbzrnh6cf7nkuOzz7ej15WvkjTkzPO5Nzu7i/l4PPnOz rT0eT86OafBiY85ujwdO0VmNS03HwFjHoszlxfvm9QvHd6nlSe1MMXT5YNGtLL5g9wtLr+JAXbMJ 2fqy+d3fL3zviITPnp/an19saZjvXbjG+sXNxZYQWYuDputnfb6x2FIXW+t434a435cnd3YXtPk7 Wj/5fW50ts7itaVNsZWNReZzveG1leUnq3oONpaWqy89aHryrpizLq9EN+6OKTrn5sc5d9DNiJfP ZnP6Sw8S7b/1sG6HpeOdVH/rKZ6vA3W2y/TdPgTqBnzWGebU88Sat5tt6Nhq/HvGDmpMwKW3o5/r Bv7zqa7IT3mZ+vlTeX1LCzX69q4u5leGZI2L6NPDHd1/AXahpoG6FP1rDO/eNculyHN6svhmMlJB /aLXhhk6O7V+6egfbtOjRLV4rqmgF+a51QnrYTDZ2ZVUTG01Idtqaa85vUwNODevxsTHULc5XQ0L 3axO/xqhi+i6tGt6t6FEINbExZCtZ+zeGbCBhkZDLfd8u/SgpaXFn03CuwcsW+5cuHJql/fkL6Bu X5nSPijiK2muBb0NU6muZT/e2ybWK/3Ej2w7o/xcBuZ4akXn59FlJwp0S+u16110L+a709UtmOOO 8KEPVyw5WGA1r8E7fc4w+2FW1s8NudSrUbpGQ1v4FY9sM+xee/q9cRX2qvXJu65/sJomSvpu+aNZ qSE2y9ycekZnmcY2ut9+zh1zbegYnbMF922rO/+og73qE6P/hNu8jfviO/TX3r1smR3PRnlk1Ya/ YxYRfxJrAydnb5ubNmHqC5icffnznuzIgH8+OTp55O3Jzjr3InNrdM7Xsu6WtW0dVKOxu65oXMPg QQ2N8xWn571ds58VGK2cTCMwhZyHmWo7p+qXD12+21pTQ69bZSZb93us816/QLJ+uoGe9T047U0t xy4v9MaZAe8MMwPKQ3+EJnjpZgR2aBSp/xJqKF9gWWt1h8jxjSJ1dqsNOy49uPSVflzvU/g7Pegn htzfSyJEt4TMLvrYsKBt6y1vu3fYVdjLTg+hnnVW3PjQ71NWpC53heHJOn9sUK3V+YEuK8pDdbHY qwHFv7zu7d7qXXd+gI3xAFeQzPX5XoYPUQd8tgjHTSf4ZcHuJQbKbnXLwXarLT+vCXw17ORdBMiu 7hp3nHQTBh3yuOtzsKA76V5S0D5RGdDO/G+zdplr1J092c0zP67c53Fnecr72bememTnRXncN/9W 1bJoeKHHk/47sirU2ivzrUkHR0er7hXtz61lWovdRRgmjhNni8tETtwnnhLTRbgvFovfxXoSLrlJ gyRfKW66tFhaL+2RBKk8MaUfGE5bJe0zv1STzZdkZsAlVy+IdD3uvWhffY8s7PGU8ZtcIdEVGnpk 2XtkRTTyyDoUnew2N7uX3VLdL60iDrboNeD4Qff3TkbbLW4+c7ye69oabdsZbdtnT2nk4jalo9G2 k9F2mtuXEF/CaOtqtHUz2nY12sb/arTtZh9itO1htO151Gjby2jrbrTtbbTtozfa9s24cV7+W34i K7JOgZoLcFaJViYeUBKLvYFJK5h/o2ThEJiXURAKizMTYp8rxiYOYGl3hrWzeX1xSoqx4VBj04J3 v0gImv1keuOng/JBaQVzK5XRYYEwq0QJVI3NGohRcUpA4otg9YBqtBv5tpT6ttVzakRWpTJaP6Ng JizIKMlT/zX/z9RmYtQu300qjt2aOjhkxCV1eEbB1GElt9/Pva/WnYP9ekevSywOxg5grB3cSV3f KDP9jCxj7zED3giH9rg7/hteNWjF41DF0ai4HdGa6E6YhpmHqNmENmgldtRdfRJnuOSa6NSg/GGq 50AJgbN5V58Sn9x1J/tYWJPNzJOVn0/CJF8SppOJM47O3kweJDWIFt6QFaQthabuodRYahaVtZTa RMVRR38Y+8DAkV6t+lhsM9/FgKMlQfU2VzzSsFm4BqUpbwbYiJZ0FZQBNK1RGX/R5+m/6Sc0KLSO WVO7HuPGoJ59OgOLmfXMHkZgUpm7DOrZQWvamzNerNazw072GHuZldhC9iNryTXmwIXrw43k+rIh 3ApuK/cXB+e5v7knnMLp+IZ8W74nD8P5CfxcfiIXzR/gk/kMHvL4f/lyHhOchF+FIYKfADOFJcJG YSmfJKQJOUKRAGVCHbGFSIgDKnnKhSKsE3eLCaIonhCeiZ/NUzg0lTpK/SRKCpB+kjdHpIvSP5Is 3RYNciO5vQzu8m8ysl9FyP+RN4/kt7Iqv5NaK92VYQqMU2YryxRO2aecUv4jb+qpuFpfGaT6qtNV WKyuV/eogpqq3lX/I286YyTWRfXCEFcHOzEE1klYIfYRs8R/kjcjcRoPwf0xxGAAgjAQg6EjNAYD hhMTiLkES0QTJhwxGIAgDMRgOJEagwEzySXkRpInk8g4IocsIqGMrEO1oAhqADWamkotpGAdpaku 2VQK+ZmqRUNTuiPdj6boADqMXkVvp+EIfZH+h5ZppLB/oNozUCWxVynsUCWx27Gt2QbMMBaqqJt9 7CkWQTdQzH5n63E458YN4rqy0zlYzGnyWip3l3vBfeGgNv8L35kneS8+iPfm1vCwkz/GX+YlvpD/ yFsKjQVwEfoIIwVaCBFWCKH8VqF/YnwlNpsWl7H/J0iXqIltcfENRKeL8FNyBaS5apLr3krJFZDm Wkc0SoSoSa6ANNcqydVKbiL7d5D7ym43u6dMkFeuZLJh8t3hUheJl72lNDlHLpKhTK6jfJMIZYAy Wpmq+MgwTV4kz1FEJV55pnxWaqnQVO2oNlMoNUANU2crGfsh5tmeskjliiq9fqKmq9kKNMLaY+7Y b1hvVXPjwSQsRj2k5mFf1McqhgMG3fFh+Dh8Nr4Mn4Ptw0/hcBhTsSI8qshAZNbFiB4EDHHpjJ/7 Nw1HjsZ1+BUCsoinRDZuQX7GncmmxO8kTCYpIoBYSh4mL5BHiHwSSkg99Z5oQw2pZDTmUIAYjdYt mraoQ5lrP3WfBIRW1qNwyo2aQi+g19Kwiz5Oa2hlKW3NfKFAQyvPvTxLDyQ3MjwD1YzGFhJn3Vhw ZHzZ6ayG4SGyErLZZ6w2mNuwJOfFgZEN59ZwOzkj7U235Jw4WM8a+NTJ9StZ+e5cMA80F8Mjmf0v 7iH/N1fBg8K1EmyEppWs/AQ+SgCWPy0gmT2Vz6L/5VuKEC40EhcJkayXePPVMnG1YG7Xjwpau/5B sBAXizBc6i3OlQ6YTz5S3C4dEuGRlClq7XoDc/9+uw6MlBfK1+kosru8Qz4lX5Xgi1xb+Sp1VkjF S0GsDkyRtytr5WuypFxUEKsDWYqL2kQZqd5RXK4tUvMUaCutMh/1ESVSLVWtsWYYvFL6Y54Yqc7H gtRw1dF83dVjaj4mqb0wB9wZ722+7pg1HozTWAzOWW/C0nDIwWcp7Ov7mANhq/YlRhHgTzxpOYfY SqzEzxNomAOF0Ka57mQ7YgI5lwSW/EGMJpPJg0QeeZW4TwNGjcV1ZBnZkZxOtaU2UsBTSVQalUNd I8uoOvQ3Egi6appDUnoCDSKdTWus9EG6I2NPw1zfuXQ7+sDkuuQ+9gSzk4ZqU0Qdxsh6s1DlC1zH 7GZusyLzPzuzfmwo4uqmTcxYuoo10mu5MPobB1+x6p2ZH7eBh738Ro7nkrg07jBfxDUToA7fgq/a mS3kjwv3rghp+3BjnL5PvKP5bphv+yj8CT/9bR4PgWMO8Q0vdTAXgSXiRpEXIUlME3kB4fKvhNZi XRFIKfznBO8nJUiiBNnSM+mOiHB5a8leai3BzHfDpQl0jLxKesHekvNlKJH1ir1ikHsp7WVUAUAr AdsqS8AQpUIBW7WV2k1trYxVhylL1WQFtBKwUUUloFhpiYErNhDzwQapi7C5al5ZsASvftzE8kqP 1D6gnlS/qQYMeuEIopuHm3AGO4ibH4F488fgn0p1TsXs8NY4HMNnEUuJTUQcweHXiPM4IBlmm+pA 9L/9mHAlgkjwIRBEl0CK5AniGXmHACTDdCX7UZ1ItH9dQcF88gh1kfqHkqnbpIH+QEIwdRt3oztS X3FxSTjN0xBFVWtvPyg3BhzNdR5hWjMqtTcRaW9X6SrtDZVxCJTC2FXmq+1LrVEFNpiF9+xTpjFX XcbBny1l47lt7GH2AvuWi+LAjtdzPXmtjC/nWB5MnMX2X4iY68349sQtrr4AKtdV2C7VMMNw/DUB oZfOQjH/ne8iAi54i4OEGmYYQXguruespCZSB6mu6GQ+bNFdChJrmGEkEQkx7eRe8gh5ojxSAlpq La+Qtkr/EQKKhAiBaHmsMktZqmxS4C9mu9KK/x8zTI4yRh2iLFL/VGPVEyrwCqOmKUNqmmEGqAsw Wh1CHcUuYbcxeKkexF61XIVVFXKoquTbsDIMdXig4mifb49XFXKoquR55oMWcYyEpoQNKW/wIaeR i8g/yVgSkFtfIJ6TpeQLohn1KwlIhA2kvMjV1A7qKHWJgttUASWRFjQipYdO+zoABvr1o4dPHEst oP4vovI99ZOodKd/o6cxDB1B/19EpR3zk6gcxxhDH6/btOjBxgT5FpvPQgmr57QcgBGcOzuPMx/l Zi6C3cLGs2fZbVwFB9q6DG3LZvHjuE08xPEn+X3cKS6du88Vcy0F0NZlaFv2p7CYPyEcThHuCLlx eL04g3W8o/71Pp9rfC7/iv/G1xV8RKhpgYoV7oiALFApfATTQfqTIcX5EtS0QH2QCkVAFqjGUn9p hPnUael/+Hekvulk0LgQW2q50pvyHTVJAe3yau2bm+qoAOrfFqszFNS+7VUSFUD9W1X75oUZ1Upf xE4sCLuoNiQy6h/DwALXHvXf8cnmNz3SfPLYIfwcnomfx/7GnmAKBm3M1VsDp5cTUcR+88nj14kH xGviGBE376RqT8Ag0pecTi4m15N7SIFMJSGReEFq4DTaoXchIYgKp9ZQO6lj1GVKogopyCKre3TU okPNHh216NCQ0dM9meHMBGYuwzLRDBxgkpkMJo/5lylnMNaJhV/ZVowfO5Ndwm5k0dISqrejLTiC G8DBaG4gu5Bbx+3mEjiRy+bgGfeZq8U35Tvy/XiKD+AhjA/ktvNH+Iv8P7zMv+fBIDQS2gvuwm/C JIERIgTYIsQLMfxN4ZHwVlAFOxHQgFpzPgU0oJ4WqudTQANq9XyaKsFd6YV0T6wt/yJ3lknZS4Yg OVxeI++Uj8mXZUkulAHFInySqmIRQhRYoWxV/lLOK38rTxRF0anQUG2r2ivD1QnqXJVVo1U4oCar GSoa68pVDHPC4FcMhbQMVZdgGzEeS8IgDcvBiswXuA7eAifwATiMxqfiC/Fp2G48wXxns3F4hn/G axFNiY5EPwI1XxBGrCK2E6tx5HyRifcEGMhGZHvSnfyNnEQyZAQJW8h48ix5iHhEviWRzRFaU/9r c4RTlGZzfEDWo3HajYZBtC89nUYcDFp/pdJwl35Bf6FrM18HOjLIAwFa97Wh0suIZPSKl0wlNgtz 0mBqMYO8+i7BNuCl/zHChZ3IGiYF209iYRW7nT3CJgSTP6V10Er5ZVaT1ldygIwRfWpI61DT34js jaAZI6qldajpb0T2Rh/kb5xTHKcvjrMqjqvn+mO/b5PiuPpJxUcMSZ6GpI/Tyw8wIRbFcZbFceXr plRiszqs9p/E6eJ9xtGnZiQFnSmdkuzwY3+4GufvWBxvObJixUldOBO88fz6Hwdjp5yvH8iWkObX 4lxJKrldPHKWCTkr3kwcXVorw3XqjcR74kuxe4aQBP2Tpoh0Ci6tFwdJQtLAJMdJ0gIxQnorPk3x 6BVafiD/eypiZvuc1kOE86UNLXOYkE5PnMUuyTPOFR9U9z1M1YNP/SRrz+Lj6r5TyV2Sdxcfduqa OO/cdUcH48U0U8PT90bD+8zk9NxC6b5nycxjKQbpqkXOzYK+6ZbdPpw3FDb7sU8/Pze1SbbqqGGh 5w81e2GY/zHJ+dXNZi9geeuiUytarz09uenLcl2/ok2Od/7QPdr6pm9KmzkhejW1a5Iu2WL9C5s/ X9pYrHpqCF3/0nLryxMvbSxXTzqnazr6lH9yh/PRhd/2d02edTeu3kl0OYaKY5O4SecSITnz/oMM eZr6j6R79f6CoXCcbcunOVcDdU/TDfXeOr6XL5fWfXbrw/fUOnPm15Khqeyq75rUv7ZRtkofL8+R bzRVo+TiuLLkmydzU/SPjEb/U2Xk1VNqUYfM5OLM5AeFuSZEo/zAFKfisnPKtoFs+dVuiuPmpMOr 856lpiKgvf/L8K3TmafPrjy68smgt9Od5p5tb2DY42C5ctK5FqdHn37xtm7Lp6bfT6ZUQNkLg9hI HCS7i7+Jk8SLEmxRZSle9CTuw1TCTVpHWOilhC6vNqvZxAjpMxHQHuzIZUSsNIz0J2aTy8hbXV/U j3W2W2Yojh8+LwyfGjp/amgT/SO7RhhoE9gkTJvA4jGrszZGuI49wF5jP7D6eIEj3hUfjI/BZzRZ EGyBIUz3D3xDy0d2+5rC7jFn8Bv4Q/wNXoHbEq0I6EYMJTSlZRPxG3OSgMZkBPOKiGdo8iazghxo btRIO3YZ8yc5jPVnZrP3mW/P+9p/6uMTGjAvTJ0TMJ8JDYNR5htRD0k0xDp1X7+NPhSEDpk3xDnA JWL54NT3xUecAzwHhff93Hzqzrzuh/JMp7JUlCfQLs7aaN9Iunq6oEV6wXvKQDeif7SnFw5wruRF y88NprX8gLKZC8fMaLK4Xf93F8viF+PrPIMW3PXrteppu+Auq5/aRnDPbCNqe54OObB59+hz2/R7 NxsObsu2PLEte/Nub925bYYdT4fNuMSzO+Y73NaDvsIQrWUYlOjI0IxzYdc/tlyW+61tSF7hGj2E ntyyeOCN+Qv36/TLDhhWHtne8MS2VZFZW05urqOL3TWj+NKFElNj4eVxi2eHnz40jhAmCteX7FKz 2Pq3wi3CcV+xTdS14OTyxP1bdnsfsj+3zXLN2xnD8vsFIZU0L/E4mfhAfH2m8Nn30wXkteCC1YRJ GZA0P/X9wa3OAc2tlwzqmFr4dPmO9WGrCgtNiT8Ki64Fa1ECMtJICwvfQP/6rf0rzEV8XsHL1BLH AH3zgArCVPiC4DKlx9I7CUBuILeRe8geMvq4L5chSt4vn5avyw/k1/IPub7i6ah0VQYrY5Ryv4J+ R69ZJ6R2CnjrVSty8MnotEI7nUeel21EkRu2YuVHtwaFBW0bvpvfqLBgZFLjMp+ofZt7RP/ifC3V q+Xoa9dCnI6GpLQKSWkdktImJKWt+T/ahaS0X7TSJSSlQ4eQlI4hKZ1CUjqHpOyN9iXe+biFpHQN Sfk1JOVIt4sRISk9QlJ6hqT0Mv+Qe0hK75CUPiEpfWHeK381VF2pblMPqxfUWyrkqyUqcv63w3ph IzC/2ve9gZkF8xcbh0CLefNQqmDIP+uxkBtHHDr6u6+zUaakhKQNrQi5qcGj6Y93xw2broNBs2Au sD/h0e2YRW08JLPBtETnAD/LxuYZ5ltI6sjEBZVLt9fOJa51TAHzYCYssJ1na8I34yGpzaYlTh7S HT9+HQ9bNTgkCB8+D6ZeewfhdnOX4w2Jh0/1OvCr3ZgYQTgGbAs/kj5OjSQghjhEnCMyicfEOwLI BiS0IXuQHuR4cg65nIwi95NwmrxOPiBfkz/I+pQj1ZWyH0yNoXQrUwwbC/wGnK+3PKKwEh2dYXpa fx31yf3kU3skYtyloOmUPK8SSk/b080Cvda70/AbPYlG/e8WOp4+S9+k4RH9llZpO6Y1050ZxoTc amF8ajtwZKun9v3eu0S8CAhJCTodGhkB8AZDEumL4C0BSdJxBq4wWcxT5hNjxTZhO7B9WRjF+rOh 7EoWTUFdXE7/P3eZVYbyq9w9Lo2FakO55ieHakP5U/4T/4yDakO55ieHakM54q8NAvQSR4gTxXmi SdwsHhTPiIDyrN6IFaKt1ErqJrUWYaw0S1oqbZLipJPSNSnXPJtK36S6ckvZVR4o+8jTZPCV/pRj 5RNyinxHfi6XymCtNFM6Kf0VBJrNV1YrEC4fVS4pt5UC5YNioTqo4Kz2Vn9XJ6sINENiFmxVMtXH 6jsVsAZYG6wHBh7YeGwOthyLwvZjp7HrGCSr2ltb9dTCDNz8xuJ78UQcQWQvccjBvuLNiS6EkfAm phALCNDyu64QSOz6RFiR0ISsRSDgwJ8MJVeS20hAatctMp9EYpc91Y6CXlR7ciI1jzJRm6mD1BkK blAPqTdUBWVLt6K70UNpGEsPo5bSm+g4+iR9jc6lAfFGCDdyZQYyPsw0BhYx0+lYptojDIg3QriR JxvIzmdXs4AQk50MIkw08xE4c72537nJXDAXycVwhzg4x2Vy59l3HPAN+DZ8Dx48+PH8HB7RhPv5 SprwAf+azzOPwsgrOljziv4hbBAQTXhV+EkT2ohl5lHYKHqLU8QFIqwVd4nHxStilvhU/CRaSQVo FG4q9pX80nGAj6CblTLvKlJIza+gn7RR4qVh6eelG3/DX9K/UrmEyU7yr5VMKcyUh0gbZV5OktPk k1KRDLlSHaWFUiV0LlRgnbJITlBEJVt5pjzIe5xlpVdXrnR4b/mxRM6Rh6pQJs9Sl6qb1Dj1pLpP yVXhlfpNrYtp81RDbBoGv2N/YrHYCSwFu4OlqqUYWOPN8AbYvKt9DIObumBTcVhgsxKfgy/CLuG3 8QL8Aw6FmAPhTPQmLAg0SZsQYrgCP0P8hdd4GhoS6GmgzG/GJrJnLThKhnQ6Qz4h37jfI+g6s0hA mrcriTTvddQi8hgFsaREpVS6TZHruDUNSPP2pJDmfYTeQf1NwxsKDU0NmbZMTyazc0pnOEkjjnA5 fYpJZ36CK8iTplnSfoIre1iB3cvcZV+wX1jaCiazxVQ3big3lpvFLeU2cRDH9aYROp7B/uDq8+Us IObMh5/GI594LH+ChxQ+iUPoeBHXXEDkOCBdNd3cUnfljwozlicLgDCkx3xj0UroI44UaRFGCRFi qBAvbhNuiklIFEEYkl7sKbUTJ0hzJVaCGeJCdrOYLjU3j8e45MLDVynYuo3cWQqUvaTV8g4Zjso7 pWOSo1wiF0r2yggZ6igucn9lpByphMiHlHMKZCrn5clKeMdPSkt1LtuThscKqY5XZyoITOGVeyqg dBCNNO6PEepsFRZggRgijRdiWZiofsJgMZZ1D8dN2C25k/moPTFwwffgAp6KcxiyIb3E4BfCBm+O Z+BG3BtfRSzAQca1iLMsXCAak1YE9CF7OHcnEJpih7WzWU7AZfIs9oz8TNaiVvZviPDxjqR/JXW2 jVpFIvoQHlP/kOi5b0cfoL5RE80PPmWiZ1LL2TO0Shym4mi4Q9VnbtE18PHZNMLH99G5jCaNwXe6 y/+Nj//G3mEvMvbh9uw/zIjl5iHNg/fljOwSThMzIY07yRZxZRwyoSLqFBB2Gs5P43bzCbzIZ/Pw jP/MP+eaCiP5ekvnxXbizYdNCM3YfQJPZAh5wr8ClAuYWMGjdAgtJwZmCZS4SUgV0SCFAkGMKBGk pYgCQWal4kWJ5fqnyT9jRD4TK8X69Uxi7QUjxQu3skV4KFZbBoZKm2SIkzkpUzolvZSby8XSdwnq yRVylWVgsXxCgRRFkC/KmbJt/RJZz/ZRoZOygQ9Tq0DDowrIanUJcFb7qYBKwBs1HtNAw0wVVOyd 2hqfWFkCPLBuOGglIB3XQMPXGOCEPW5q2ljhmgcTJgdPFxiBj8P34UdxNDCjeRnakw7Eb2Rv4ndC IOYTNfoD78pcoO4UDKN6kLMpf3IjxVP98MskvMhfQd0nbejvJIqK8KFhGu1L/Ukvpo7TV+g86rl5 aKZeUF+oDswvtBYVAauZcPoos5NuzISTD5g7NPDktN3N2aH/IUmAmKQDzAM2w1zmkWwCSDf5lZ3G IdmkMhdGq+NfudpKM76T+WbDD7oLPoZvzO/g13GVuTBaHW8i1OL38L8L/XhAukmM0FM4JxzhURkH VMfbCu3E9sKxv66SfV5CqnCbDBOSxUNCnthfUUWwk1pLiBvsIc6WxotzRJAqzcc50nWxTHot4jK4 yYNkX3m6PEZaL3cVEmRAiR+npBvSS6lAtpFJBbyUICVcWaMskI8pu+R/zEepvFcMKjIfZyodlL4K hKgr1K3qX+o25W/1idrLfJSqHdYa646N4vpjNwUWg2jsAJaMZWB52FKsHHujAo674YNwX3wo5oDP wt5gkISn4Tl4EV6Gv8JaEFY4vKG9iCAinNiI/0Zw/TtpaytUyasKOfQj4gjNWrydQJEhUFXJqwo5 mMwHPY08Y37Ck6jVx79ToOkimiwyhlpPwx56A5VKP6SuUveo2sxXCjozJOPFBDHhzBpmAX2MgcuM xLxdwAcUT3/OlMxRaWjB9GMpNoBFyjaSQWAH8w8rs+/ZAuYDU49zYMGZrRkYcNNc0VkUGGDHA1cU ueJI9x3AJvioYfwqvkoBAySBFXCaAoZcJPA7P5nvJ8QLZwVNAYNKCUzQFDDkIoE5AnKRpItVhIYJ SWCOoqaALUvBO5xONAjJjtGpPkgwA43o+C5pcOkgc+tmPvVB0mJ5jyzI00VnDrJlN+6l9IuibbGC zK2b+dR3Kmvly4qkXJFfyPBeQafeRx2p0io6dECnfl49rDxRFVWHnaWhLVZI/5hs8V/7Bqh/e4Bl qFr7hmGA+jckl1W1b4D6t694Eaa1bzYEBOIXyHldvIljhPamQ1VuZRNCi60EfwKFo24j/ia12Eqo yq3sRWqxleBNDds0Dr9IvqY0Pzl0/bmp8KM0PzlcpbUtlw2j+clBM5SvZRbSxxnUokPNHh216FDV o19gjzD5bAkL1fK2lnILm7mD3BnuBveQu8kiMwhUy9tL+U18HA8n+Sq9GsnbrgJUy9uxwgkhRYA7 wnOhVLAWm4m1hf6ipwiB4nxxtbhDPCpeEm+LBSJ8EC0kB8lZ6i39LvURgyWIlGKkQ9I56edkKkPN yXSCFCXD/06mjgpok+kM5Q9lg7JETlTgqnJPeal8VWzU5moX1aiCtzpFXaCuVXepx9XdSmWSzycV JflU2Q4hFKu2HV5USzBAewqU+lC1r4LN+EG8WgJ5i0ErAkkgY4mqfRWcJK4RucQr4htRl2xJ1iNg IFm9yUCLDLhDok2GlvnSn+pMAsokrtpkoEUGfKC0TQbKfJlMj6Qgko6hUcxHJo1iPoCBBkwbpgfj wYxn5jDIzgMsfZq5zjxgXjM/GGTnga7sYHYMO4P9g0WJToks8IEa8F/OoubrV65gCOfHzeRQ2QbQ kRo8GsVd4Jrf4rJMNqA/MCuLM/CWBosoewtei4rwiYKX3EJ+HV9VyqfwP00AGjn+lP+PHL8oaOS4 QYRGYntR07H+hxx/JGrkeGvJWIWCI3J8TlRKjj4qxSoqxfXAVd8mEJWSnBp182jq0dRdaw6mOUTn WESlWEalrKsTawC+Eh7NvRaVavT8loZl2+TuWelw4Gr4oZRH/o5R6ZYVHe7owq85RHXLW3/gWp28 5FC2pClpfi4elJD9Zc/7DtFLJ8vBWZXwKJMFS+SN8uP7Y7JfpNaRrQquy9DV/PEak12caqG0kJ2V GDlzXYFHVEr9A9f2F+qgj8/gp9f04Bx0PP2FrUP03beSlJM+I+raodSIwsZ68ElOtY66dSh1fE5m TvruqBtOeSn0S0eHi0ZjrTsma2Pg8/f3QnLmlBYu+OK5fV2KocOCpxZLiwue5VjuzDv6/p9mB1L1 85cVioWOUaImkqZ+NhyZvytb+noz9TNkfjlFV4qkKaW651/YoZqF/lnW1DYNo/WHCvNSdRZdLRNs 3KxsLDrpDV2tLCf2tfK1srFcbfGg6bc0qJWTfbOXbt/VvHS7j6evNledu6hvJTWVs3iQOPH18q/G eWo9fL5y+OuOh0ff26Z/cFj6NFC37oXhdPkOdYqD9amPYQ32F2JVOwVHRdGVOUR30jX6VLcWfzzX p/bnvuX1R5YZgz6nBJXefBdyP/B5fsj9dd/WGs53ntmwCyJElxdYsD8J0TNOT+pcbGcw2Y0rqGh1 8kCq1TXn086mBvedGh9e+OyC0+lWBtHJ6kdrXcrNVtda1eu2CT+q+jc16Gecs+th8OtqmNTDco/F g4SWI8tMFiFd6oa015XD1/LxhP/uu77ltzHMEf/yxPtVGeWsu1beUbqWscxw9t4jNa9XJ/3ZBj30 LYa1mtjTqpOhJeU6Sjepx9Rhnm4j5/RkKHfjKr8Hrb+l/WeKLyb3rAh8aXieYgCdzjMvJc4Sb3/6 qvl975iXvvkR5rvZaMjrMFGJXamHvrj3E3JCpPNYuuDu/sIZDeAyLuGFOHIANiZcCD39ZiRBd/a3 mvLJVEZ2NH7N8s5b0SVzK/EXAeeJv4knhELoyIZkW7InaTWcNP9N7QLcxyxYQA72jU6J7h4SkDW6 1qzpJ8h6S63110lwPN634UPX8juFpKY0G1Kzv5N3FjUZNGaBwxrfhdZ9BjfmXDbpxiyYvKbDxqkL YQYXT52lbg63j0bZcQRjqspAjqFn0j1jH9ZFqxUAyR+Neg9oE/IoGNRG6lXMgwno6Hg0uJ1+yuev K7q82jSjCdl5sM93p7JLEQcZ0Hao1aE0MLLf0egA99/OMKzviZCAscfrHzf/Ydm3oR8X3GAfspBw 2UA1GvrbmToZI872GVxX0re8rfvtzOgM/J/xZydJ7kZA69LvXD0e5+fxg3hf3nE6/z0j+LCtXHB3 1Fp4GQcFV1HpWXvHdd+2Jz2O9ZkKrvtqxy/NjsxuenCG3draS7Ojlmbn3SzbaizrPvLT2juk2txo HLy43ovu5Zdiw6GUNH5oRmfOdsY+N2r0steLVS/Wlpo2q23FM+qNsuvi/mbTag1LKRYd2aarXvx9 1EX6rKqOKLLiwcvnFcmi47u23PIkblVhYFaXzvqZ2T0MM7Nre4QPnJmdkKqbcafRB/2Xp0MM5S8t a5V2+NDog86udL1hx5sbl75K4Ud7vtJ/B/3Ewse9X6cmqAfyyVaDc1oMNPg8cbft03Buw/EPFuFH vOqbdHoIrfPRfr/TvawPlrrcj4YnZX6/1ipN9Hb56FDngy7W5UGx/7fhBcGqtXqrr/WOWUa0xcNz Kbn5KHyvjUU4fgJbVtAiK9FdtCud1s2u1HJNrxnD8v0/IjW5q+1xMtHcrzoU3PWxLyBbZBWsXpul DOh9MMnrVNzXbc2tHQ6stCu4067rsXrxBdl58KPgfousmmpywWPo378gpUKtDVbdjyvpFolew6ZW rM0qyFkrI3wTauKbp4h0Aqpi0HDSjURbXsfqNW95ouvR6DZWTlZOlgVZT+1qtThY532rD3FwLq5x g0V29QuyYgZAkldmdy0+i6RQfBb87xb4IwX/uwVeQZ9DW+BEXwTHes5/Zx+H4FjPUFjcK7PFc7rg AYJj27nuy7HpPSWl4O7QgjzjAd5inWeT9E8lbzJ032FQmmclHMsh6beEDmQKHjYwXkxcmp3oG8wc YAqyR7ofKqdyTVPBDVxh4SgIyPBEdKy/NdorF2Q3u5i4K30Tg7Pet3KmHEkV5wzP8Jw6zF9/O2Du ffNQNS/pYmLiuGAWDrCu++649Y3P7G5FoZ4btMC16pAuqGq6tZAuc8/dmu/OD+PH/RfSZZO4W/fE 1vAhLjG6wdyHbqQm/+Z1fcpjn9xPhhxk7YV2gvAQEv5BYVyJD7a30NKcIFOoSnPSfFtGNFFpaU7t //A/krJ94Mihfeb3yL73pbCTfferfcJsj6RcK3V9k2TuaS2lxpKL1EdCiEGIZO0E8yWkpFyWJAkl ef0hQVWUFy2HyCvMkxf8JVeTzA3NkxfUJJkPoNinDCVP+VcpVzDVydyDD1EBBbhp+W1J5jEsR4Ui tUytg7XAkDf3hjoVAxTgpuW3ZWNjsc8Y1MKb4h3xfjiF52L++Irm6uX4mA1h48OC2BGzvd2DRrjt w02n8HT8dsMV+DTcQPz5i6oq+fm3r+dfT1bzLyfkX44vShZii5JjerTdQDoqjxJWDmkTv2RU6y/x y8L8nGJnf4kN8m+diJjUDURDQilKdM6fdalDUWa73Hsd8m8/c3GdZh7d/yRiifjYxI4nCchwySXe EevJBuaPeg8ym7Q+3yGAXNxtdgS5hXSlzpJ/f1F8RipTjXn9chPqbVgyIH5lxNAn7WyovyiLLpSR ShhNjfv9MZnRbka3FD9H9JH3oCXq7cCi63JC0fX4IjfHy/F0/uXY/GTT6Jg1ow25CZznsoFLB5sG fOjRirYV6aH0WDp/wJevdIX3O8pkmHyr1ShHlA3ozzCzr+vzzY1xQKj3hqCF9GqfqImQ3q43c18z KgXZsSptasxcaHudCWIrG8twBbudXHT9YDBasG/2KRjVY4gjYh4tNiUEuf2bML6JyaW5Kd7b6UQr U+yIJs6mmE6mgaLthoHWK917RLoZYEn/yOZsbzafXcLpuStprrMHKgkx8SuKlIQNRfGJEfnxK8X8 2JW5sUsum273iQlLW3E9JsxoWuQH7IQoz6ChpONcrtIBZYpIfmMatX76gKmDgz2p6CJdUZqpn6n5 Vo8N9IhSbgsf24yPixihJH+hB+5U6lJbxka7W59Yxo61ndbD8RWaUteyYZc2q18SVG+nDSNcrl9O 6Ns2jIrzdrdNzjfmbd2UZtrgbt3t2DTe5kBs4s9ErLjL8TYhS4jx+WEDveOUkVH7ggcFuXvHBcS1 dXy0Wblsat7NlHLgdvxKa7HetF3CuZ7iFcEqcYn3l9z9t+WVs4tiwvIP5t5emL/EO/9A7IawQVvG /9hvu/zg7YRFnooQbO/m1MsxttmFy85ib/GaKOUuGf9lTNyrxA1hYvKSoNzYlfkHrveZPWJ6YnKQ e8mhy6bfYkZsPJsvmHo6Xu7eOiX+DBZwUqSkXPFV4srxF79cV5LyY5aUJhZ9vJwwPTE/buVAp6Lm 0Z7y/vgNS6YZz5qcTUM3nVGu7zzd+qrbrfgB0o/RUsnt/JVBX64nr16aGr+hNNE3qk5WdMqGniOc ipwdLydYpxXFLYhL9XbadOZLUW5YonHNjGWev99HxPh7aYNF9jk2xdycpCQsCQq7Ort9zAiXFuaf vm51tSzRNsnUyrTpTMNQNSzxiXGHbcrdu7ttdzsslefJ1v8WbeBmf7l9mU35mPgx8XZYmHdnR+uC WG/3zUl4mPftJeP7FOhTzuV/5UP6ZD0xKrmXldsNLxcdCNmWmPBCZhUrJfLvrJjmhTHK0r/ahZ8w xaeNd+l6/8mLiIALrb7kY7lfcq/PU6xVTtn3XBd2RDmntA2ONzU3DVJfFUd9UaLVJmqHb9dbFFz7 dN0y+Z8PCW8TAxJz4i5HyR96/RXR1HRcvZ4cf771Xtu4d2oL7KDa4esIy9nNCooMG4I2ZemzFqS8 TpycODTOcWjcftN+k4OpF3Zbp99h6/gHdgjrjGViT89crpX9GoMf2IrabBJKnjuL+eCPbPaHNEmQ skz4arvk7Q3jv+5tSOOJDloUcpptRoOfGmYi7oGjIIGqFRBMIf7Fa66AYCbxkajynKIVEJQR28ia KyBY17Pas4gsi1DTs4hyo6GmZxFZFn9cojMaajbCwqlplcvANFvzJOi8ubfBPG4a0mztezI0AyHM CibNtilzlLnE3GYKGPjAWLDbGWfz1f6dnYywyEi2EXuIPcdmso9ZlCsJKFiyB6e50JcjLLJmmgwK k4GaaTJ7+UQeaqbJVK4EaqbJXBGyBKiZJoPCZKBmmky+WCJCzTQZFCYDNdNkKiRbGWqmyaAwGaiZ JtNSqREZjdJkfkZGV6fJ9K8ZGf1flEDNNJnJWDAGNdNkKqMEaqbJIDcy1EyTQWEyUDNNBrmRoWaS PAqTgZppMsiDDk/J6jQZFCYDNdNkkAcdtHWAliaDwmSgZpoMCpOBVkz1OgCFyUDNNBkUJgNaVLS2 DkBhMlAzTQaFyYAWFa2lyaAwGaiZJoPCZCCS1zB3lCaDwmSgZpoMkgFhv6Bh7ihNBoXJQM00GeRs gKsiYpq0NBkUJgM102SQswGeSp8kZG1AzoZRsr8MoTLKsjssX5AR7lwiA8Kd/3+/jwAQ7jxUHatW bfjgpHpNzVW1DV9LzBUDhDtPwxb9t+GDO9hzrBRDG75OeH/cEzcF4vPx1fgO/Ch+CU9rOOdomq0+ zdYqzdY1t4FvRJM021d2aU2f2D2xS2ukO5prb75mFmm2lmm2636kNbF3quLAardJszN2bj3QqV+7 Wm2SHXIbhOfZ+jumNbIcWTGplS7c3nG6y/pc+2n9iVLSm5jiXEKGECUr2ts7RRMHHD80y5xt/jh+ Jk8QKQTWc50TNHF6Tri7lhFTiRbkOqfmTnE9SU9iAplBXHL1SLNteKlRmxw3hIE1amPoCM5ptmop qdo7WY08rO/4a/e6TjMsnZvnNvCJe2X3a3cLl95t0+x2pzUp68H1tjVu/TXO+s/B9q6vt3ZTDrYt ONhWJj3H6FxSDEe6WgjuBQ7dZ1recskb0CzXTj8/0a3ig5tj2psC61YJ/5wtGWKY/49TxcBJL2gU qni9ZAiUjfx8v2SIrqEf+qoVvz69njq42to56fPcOtADLSZ4GwLGWObY1XUy1rAIOftzZMo+8hR5 q/ME7wddr3rZgM0Eb133abp1xDvf3UNVGuzILMKX2TliPrOaaUseZRxLx6QzdswCMq1hVz173Pcz ltsg9KT5B6HiDyZUFx6GohQ9Z5S1HTIorZFnQGjAdDU0IHhGWCdHHyUgcM7UpGDPmUHs/PDQqWEw 86yPs5HyptKaxKQ1aX3XJ62Zatseuoaaxy6heUYqecLblN/N3iLXvrCv/j1pHoKppDGuI20ueLW5 s7BuWvM7o9Ma59k9P+nFDK9/0nNsyLT5p9WGc6Ch0fhgp2K5wvvKZuOyhlP+cvjFKmGo6h/+PTKr YtZ0z4lnkQ334hJR8Fpr6PxpYHjpcZN/+B/B/hbPA+lKmaVbdB33bWcNJjvVv6JVrp1Vp4E7uu4w NfCNbtwirFd0h66bDf2irf7Yohu2eeBmWB+DouXPbp0wukfE1HGPtv1ruaenc8vClc+tZqs5daCB 0diikzL584XUzcZXheWq1y2wNxpjaqkJtq8vWBvN/1jtJ22Ov3vr/v6HFf1px42/SZir1w8/r48j Zw6Rn+3YWf7EhnZkrz2yaVv7dz+vfRGG7pH0oEfbHm53stq2/+H2eOFsvO7Rtmd/RQkJr7e/Fc7G rfKcYb3Slf2pxThH/gu6CQsMlz0rxZiTnkt896zZ6H2cu7L+pO+267SvrlvHDr7scR89NGdXdb6c XHcL7MgV84LMn+hGPbnh3ATuZxvJfV0qZnB5GwYOWjL5kNq3qdG4t8fV5oaz379w1dYsCOLD+TX8 T2vWTiK1ZOq4e7mf+KuJKTlujurQvk2Mxqc3mn3hcHWz8VritVjVql8zo3HcwZlfjo762kEAR6tb smXt0HeFQwTHEvnfxWtil082/yr7fvf6mDszny0WWMenU1ZHRi/e/JfxXm5M0d375w9uVf56p7uX C4+LDr99dv+lUoUKw2xRkKq8rPZSsWj9Xbyp3p05Ks22vu6CV08JhkvD9Hl2vvUivx+ammfXawhc 8Rqgy5k/wSut6RWvx4tvhj/Ps2OGdy9c+SPIyWgkZyGSrmCzcZh+GbvbWKmEftZZlxgSjGWGgrND u0Rco5LkH6+ob0OruLpQ5b/v5WlDOw5iNyd9Oby6uL/iOLLt2saNT0738+oy0bu+burogfV1tWtb zXnfLOxeA/2ZBnFODcJHX3dwanCvgS7LwZC+o6fzpbPq9ykl9fVQru8SoXF1rUvIMsuTnvX1eonB vDb3WtY0OCyzIbRfxLXS6be2NsS6TDD/Jv2DnBpkNqiji/WcUYy30TsQREUni7Ib50YYUXivyxKn P7o5WETPa2WPzyX72rYPvipnORyf5ZPlYLnmwIxhd0d3rd8qMuQ4apOGbhndcPDsArJ9cMEwUAYc 9zFN8OqbleR9xSt7fnbIFu/lBQ/Zsqk7XLf4tUc8nfrTc1y4YMtE6H/es0KtHT/ZumTBBM/09Iph sMV3GPYUBTlpuc4o1tmfDqWhKte5yoRgrMp1RrHO5Yz/vn5XB6cP+D1k34BafkZriwz7P4JhQNbs Dz0HbAxhQrrt1Y3MAiYLGv66kkiNm+o2P3uXjnMK3PfL7fktrCuiZw9uH+LkZ/TOto7YlXp0Z0g7 P2P7KAuXkk8d9/sZO/kZ22afsSP8jK5H/YxufsaufsZf/YzdfO7q/Iw9/Iw9/Yy9/Iy+7n7G3n7G Pn7GvqtnQHVt10q7qbq2B3LMOu9/gYk0zY91HALzNq02hcLidSMPcH4eDqPBMmvo5QE2f1ik+A0q H+o3vCqL8ZTjsS86GBRpqpHFKHBOvN+IBstdT+iYVQTvE8j7DRi58DrlUzLVGKxlMa42oSzG1VEo m9pvQLPlruM+jx3J5/PrLQeHLOGHrzatm2oVtTF67ja+o/AhUK/TMesIIVC4PCA64I/0dVMhnK0m iBBABM3FaoIIAURQkyBCAJFnX2mU5C8xf+tijYYlWV+Z1OyNOwN+ZjHuCNwnfXI/+S0mQ4I86V/J 4dOccnu5nRzvMbMOoBCymij+/+cbCnu/jDV+ixo4stW3GN+QzAG6gJdRQQ4f0PcAa1mMyf7HCm9S cE3JVV4p3/7LsjJ/5lCWlWbASVGHHoFb/8M2TsRgHmbCNmMHsTPYDewh2j0/UrUv3qv63j3Qvnjv Gp6Lv8K/4XWJn19R4UNMI1A2QSwBqDG4QzwnSv9Pe98B18T5//9cAmE4whDUSr2AewYHbr2AIOIK G6qtYYigqEGGC9skTK1oQHGAaEBFsRUDKI4qJlAFJ8GFWpEEUaKlmuAAJ/e/J5dIrFr7/X77bV/f /48Pxtw9ee6e5+bz3HPvwSTZ1KAHU1/F8iQG3lexnMEG4zF9FcsaNoAqlhac3pxRnKmc2ZxFHPAd Z7FW7YYUuwFQ7abNqiCFC6BVgb7jEIBqN22OQwd44BQvn6vvOAR0kHzSquCiEECrAn3HIaCD5OsQ +aBV9Ehopuc4BHSQfNJxqKMEQLUbVNzmOARItZs2xyEA1W4g30ZHtwE6tRud4xCAajeQb6Oj2wCd 2o1OhhVAtRuevE2GFejUbvQk66HaTaFK5zgEdGo3epL1UO3mMq5zHAI6tRud4xAg+ZFNdJ3jEGhT uyEdhwCpdmPO1DkOgTa1G9JxSEo+OQ7DvNlcF1KLMY6V5KTVYuSy4TsEjstlNlBMl2iQ1/04kE71 DQdAPOk3bBJPeptzng1wzmM2iSf9ihvO/ZYLhNwcbirnDPcG9yH3NVf9Vcdn7l/wEhKGLv0yypj7 iAPm8sy5cbx07ePjcS7Q+RKQPouDhOA74Tw9MdJbPK0Y6UBhkpNX/NcZjkLwrWjD91miBFGqkHyE BKbil0JSkNRG7CUKEYFt4iOi7aJKsUREPkIC+Aw5SAwFSUOJh8u9EsBOLpOsPVApeS4xOPebePk6 ECeBLyl18lapEvCz7LBEKbupUbCDmvdgvDxUvlIeJoMIhFJ5gQzUyw3kRqruqkEqaDRzNx/czD8v h3ScLXL46uyiHMCnkF64A+6Kf4WH41+rAHwKgXScIyqSjgOWJy3DETn5HAkfIwF8jvSUQ1lcOQ5f T2tJ1eRzJHyMBPA58hwdQvRxOnw9DfyZuMocv4Mq0GloGZPfCi4xW5gm2DOUifXUWJKA5dh8Zgb2 PbOEeESow85h4CbWlT2A/QXGZk/CotiJbLCVzceE+A/YHfZI9jX2aDZwRM04a9ZM0IjaL2Dv5GjY khCDMIZjzH3JHswFQZxeXEeOLzeEk8ldyznFBRAnfJmzjLv+x468UbxEHLDlz7h+vEiegHuEd5Z3 lAt+47XyzIQkrd5X6MoD8bwNwlVCSKsXCh8Ib/BAR1Ga8IH3aNE24T2Os/BbEQgTOmqUEW6J9gmh MAJoFY4QQ5yCXINT2EFc7qIWEVRGeCl+IKoQM4nLXewlmZU3Uwz11wcIJ30PtoivS2RCkneVdWqQ DHjLJktWyEIlKyUHZTskYgkolTyT3ZeQ7jMnZBZyEC0Plm2TC2Rb8Ep5V4lUBo7Jf5X1Vd2Tz1DN VUWo4lQgXRUvP6aCAMLHKjjsDeC49xR8tGoR/h2eiu/CwWG8HA/Gf8WvqgY/H4zfV4Gg1guyAHQx 3R+HzIyjdEBiySAtlkQTA4glm01fj35Hh6PYN1BAita9oQ9jhqB93iTZAGcUV7kyh+PHmUfFeswr CCT7BpvBnMsEEcw4Zii2l3kLg0BzczafZF5BINm37DhH9BVuT3nloNVi7CTJwvqu34bZNodgV2aA h0QjQOISIjhxnLlsAHEJ+9h32RfYrZyRHIQDLDl9ONZaXMIuKFcIcQkVnKucu5z+G4y5X+AA4kb3 oOt4EDZayrvKA2XcFt57pnM+mibAQFgilAnrhE+F4C6vq4gmHC+K1pnOzRCRTcAdkUoEnaYARTRa PFi0LYPJ3Ze1RrxtCwjbHyRaKjouKhO1iE0kPSQAGpQMFQdL4DhShThFDHQDSXfFCyXwZQaAbzNm SeJlKyRQ18pHBK5L3s7dLlNJ7DReU7PlABIjFst2ydNkcJSoSQbeyEk4sZNqhByiT0CKaqdqvbxM VShnqtZL1Crwq/yo5LseI/G5OFsFKQqA5CicUKlxOdHMU3FghffDx+LT8e/oXFyAH6EDsh2Hzbgt FyLEwXdoF9UU0RKUiRagOfRqFJDtOGzGUbQYnccEPmgouhI9wGQzq5ilaDMTkO34ROYkDGP+3O8X CfCaf4vZIFnHvISdYjZhvlwAL/Dx7JlsDhvK1kViIAFTaoYQH7FrMXMOhQ1Gc9w4cziLOTzOEvZu zjQmIFGhF9gKdiv7BceOA/y4C7ixXKiHsIHzM7eIA3RihRBFfpfrxCWai7W87TzSaQrKIQB3no4w PZ/uK6xjAtJq6pIQ2lOkC61EwOAdYXqucKgoTggMtP4U0J4CiBzEoKfIQEXqnuSKgsX7TgFncZts HUoSZ33Ex8QkRhwa1YCr4jbZunfE2RWS7ySVxC38nGznYAAlZ0fL3eRz5IvlPPkSGYAqw3tkt+RP NMOFtipgJtepDK9XQZFhQKoMG9KO1iEtb1TGb0BXlYOKJM+S2iegFC9QkeRZCEXqQ7ToOAsnybOk 9gmArmJQ/GQA2o2OU7YPADO7bR2yiLdOixG+igLSq+AFfQgTY3owWSjwROehy1Afps4GGbyPJwtg AhJPlsfU2SCD9/Fkm1k+6EV7aoUDY7+jdzwGdLbJ8NqFBx2QR30OO41TzKng8DDAoj/kuNFb2SO4 Llx40AF51Au52RwoglHNAW85JpqjrvMXA9u11sdSrTolkMknCl/Kuzy20XbfgE6sUM4ju2+gnxAy KuGbT7L7BnSMSlyoBQLbiVeJrkiSDi78UO3GXqxVu9GBzz5Qu3GXaNVuFso4nZeKrkooctiBByRJ fomcL+fK9siPyMFZ+S9yUqjUTjVSNUUFdCT5bJVQDrvoQL+PDrvooK2PfgUvVTXjAGpgDKZDCYwg ejQ9iQ620UmSvIL+hF6HW6NAny6dju5FAaRL16CPUcC0YCLoKCbQp0sfZpYzAaRLv2F2xhjYcMyW 6YuBMGwVloLtxHSOQwBaDkHHITgINY/thYE17Ez2AfYpts5xCOg/miZwotifdgXlczdy93A3ccD7 4I4pPAAJ8qTjUBHvNO8QF+jQ3jrHIQAth0jHoSvCe0LiJg8th/QdhwB8rCMdh0htUWAoGifWdxwC 0HJI/x016CNue0cNHYeAzgNXZ4ELXCSkB67OcQiQHrik49A8+TI5CJFlyvUdh0A3DeqXdBxKUG1R gTzVVrm+4xCAlkM6x6E9+BEcnMWP3pXjTTiVTna+pPqEyjYtxjz6GDgY/ACYAsqJXx/QTXugVBuz LjYkvxIFi8xa6UIUytyQqHCwmmjKUeYw5mQmVJd+hYJ1TCgvXcq8yryv6bJppA8wzEOrRJqBAZ0S 6VMMSmAMYPPJW/pSNqlEuuiHPBYlj0XLY9mfcPJBuuWxLjnmuZU5ljkWGZ2c/MNQc4M8lmEea20v G50W42PnPMccloVXPw8732ID6xNOMUdOsQIYeS6GrU6eyNqYoWYzvlp3wrnXV5eeQC1G4nbhr8Z8 OWF+L4aaL+OscSe1GJPdIa3smd8Sj7eOvTg9A0AtZxpxei3xQDxsuA4cFjf9ACcnYGoeq+sJ558C NVqMr5wpoH/sYMadkKHmjYuV7EcuYXnOpxyHZgRSgPclR+O8Gacc0yN9Hrlsz5tq28RaPt/shC1L 4CUxZq2a13RtrU/CsroNy93FJlLqwA1BBukLFK99DAu/Klt66YsTjpTozYE3At+h6W7FUPdHF3ko V5y7FQPuLl8tqVk+PQd7eCsGaX2zvCeEjLHvnn/t3nuQOeVUYJMjYjB7WqKpW5KpgXMcdVrScUPv pMVJpoZJNv7dgYUX6vPQzV1w3KnJZUDURSd3eH0Ycrp6CG38C6IX8resSOL1EaVwpSsKZpctNRt8 JzI9KAjJCaFe/LaAB1ZvuRC1buNPgZcch5rDfhJQbWxyfJuBc8/4dOcN4t0Unmbz8lgvoxcsXdq/ C4t1BPV6iQm8cOVan4ema31uZi/j/8A7yavkKVbySNJx2VMeTkLBSktfQPmpt8v6d1+3cFuPL2u+ TadSLjnOivctGkL9+ohhgo3/lxbee9YP6Xgnkj9jaUDrscQyNrjGpvBesKFh6Cqup2gnF8zj0GVb snvL1JypslLOokrkgYcoVWbCPSxjVFTJJshYc7g0OSqD71jWXek85+xLiwm781ymcCMXB9Kjo0Mi b2zcJijjNohAA9dUbMobuneoubMA25NZWGX3pdnDU47e33YOiT7ZBT8ZXYH3ZLEC5PJ+vKA7sRGJ lnKN3GRMBevbV+l3q2Iq3D3FWtOdo8ga6pAawabMgoAY07iqNtOdug4HrzRQ+ea7Cu1OONJG3n5w 4QG/qIrVeiq46u7AC3ep1XU0y6eI4u7tu/a9n8GBoRUZVMo2QcDx6uhmwx3GZ3uGRI/dh0eOCsW/ ZLF2y5fym84rYiPQUjzuliVgsFg22ExZx1e3jIXZdiW8ZbaXa2qHNYyV8BhGOauzBoWxBK013wY1 +YZh+cpWQwnjVfe1WxnBDFbS1rhZcSONolpozvEOdIprZyS6ebXR1E7xLUl0d/A8cc7ZXq2HINRQ pBGbjLo6v5ra80EBFJu8U3AscZQy4FgR0fFovFPcJPFBDC+YDAyQR16jrJL7XvK1yq03znjARe/9 PCuPVbP0mui+HCitTzk+/9JKPCKXmfuwsM4cYHv4aXluDwsNrr9OvFPsv/KtQ8hKRm+Ih5N3uu9w 0mh1TDOGs+Bbov7056YxZxq29rlfJIbjEg/ErWNyykk4XKsFnUHC4d7KUAmj8uFA2enXFY9qcQ0c rmuhsE4pHeoSV7N0Vjzxn0kKpWZpZuGZp9HDKd/7Dqd+Pcpw/qi1w6OHI9YRo6jbjM+e9EMX/LKl fwuFuDfaQjQcnq9Fwxk5ed/pXbXyauIm+aXNmQkNpbKrkQtHjNu19prsCkJxmEyd4OZ7aP6oiXZr Riwc3gHJmnP2oWRcHkOGzbUx2D41ZxUL0iaSb38lSpN3Xn8dGvJwahiJVUG4f8SokMMRowyTD551 bT3EhVi4NPcDWCrHf5dvdvwhbwWWWKVQmk1WOUZe46fVKGqO2EzYbVDvkR3X1//HEbnZgjHu2cmt iVosnIILR4KzU8CkewXJrbhJ1SlREj815H4rUJplJypRUuoXKv0CfalfaLcM9KV+ITvWANJj39L5 jxTzgkICHxXaHSw0mrxn4bAealCnyCgyuOGx9zJ9fO8BiSRzFuioszorZtBGnSWZs1IddZYmpud6 ghiuXXb0OIiC65vtCr1zFk8WDwiz/q3gfFN304LT0hPXBizQQeC8WMYaCJxddhsEjk10oQcstGD1 dH9YSM+ViSmSAYEzl1bfYT8DIWBYYcEs8GVg3+wFYFnfAkjDjRsQ+EVP95eGj8TfSQb1mZdectj2 uFvf7JC92UMKF4NRkguSrU093en7ZRKK7JtA4Hc9MCRQCOH/0Ow3Tgbeh/8TT9v6xHTIS2fpw/+R AH4dtYeaXrc0f6zA3SlSGRMVNibOUfVsbBGylCqLUiWqOg/s+KVGZprev/mbpypDvCs+AIfED0Ay P0jiB6T7PIDMjwELA8YxnGZO2Zh5eJR9b99hg+dSMg+ZUcYxTvvwpyMLuOSAHNDJT4cGXqW/oJui wAYdquHw3aIvQ9egmSg4gJ5Cq9C76AKUxuzGHMgEUNUM0j1eorB7/hMTXGTWMtVMCraW2Rcbg03D wNfYEoyPGf1ewXIrplOwBKSCZT/2aY163Ss2II2Bj7K9OfM5KzjfcxRZnIMcB46Ec/frktx0XIN/ mxbuSYf4N8CF5IEhweq7nDOcIO7DORD+hmvwb7Ul+bgG/5alLI5J3xySsjNAlZ+wZ96J3Nj84Jbc qCOBWXSIf5MET+DO4jZyfSO5KiV9RW2PVcoByzpVj1hVW+m08gw3iPeQS3vNzc2ix5rz+q504M3k gV95S4kTfDPPXthtlZR3MzowHPKFhUJrYYvq0PrjqtP8Meur81N+EeQKEhRJE5YtENI04mj5oEh4 LGW8sO+yiugjAfBcf7FPNFQ0Y6OyPF9Z/jpXmRpQkltbkhVeWyzflN6wqTr/cSqoibudIBd4JAlE PcVQ5Z4yMa3FT/zNrpmioJ2DIvMD3KFCi0Rc+WP5xFSckwMub2zcXS26n67K6r0sQ0x0M0vEMnHd 7gjJXPEycfeYPhJ1mcRNMkeyuFhVWaxEyqnSRvFbcVN6/rI9AYclq8slwZX5wcPyZy+X3hkQK831 tBVIs6Y975YsTV8ndTJLcVpunDB2c9kwELujTBoryZBMJJ4nA2V2/HCnFlV+usFVZX6K8m0u/XJt bkJtVsLy6qxYm6rK9Ci7q0bl6VEi6Y2ttdtVqWVJjJ0BF2W1MrWMUlU8vSr/gn254HSCLPXgZaXS ripLyo+9vu/iqe995E/kK+V46bRNqmJfcc4t1cGrRy+PRYw7Xardxt27OQCy2pWSt1E97uIt+Z62 KSIN/i3h0EGFZ8VYs+KJqWPu21WljDU33tR4Rr6gPgsOZaWrmvcqSnIXFMbOro0KcSpUqA5UAWnw 2ELFGqkiIWDcXZVN1SYp+rSSmpsQii/a+1K1Bf8SD1G3xHq2ODypTAhXvk6PmvisurI21nPWxKdZ KVG7rs55Ynbn3rrK/Bupqs7S1MCtAVl3V74qScYz8F706tjZL1uKFa4gJao4NpjYIQkTn5aHTytX hxYHj2U321SlT/utxaS2c9WWgJL0YAPakhxz+oqDdAe6K0iY3VI+C1lemx7rA5Ql+eXqnRPVEP+m Tp30JPfnlNgzfOuqZOmjZlX5mhfP4iipHXOh0w+7Z61cg3+7jeSmqH3A4aqFZmqQMs1WWZ4cUJIf Sp2ixhFP286PmluU1VfUDXk1qT92NSbhHSnB5rVATfy15McGX6GEp08bYLCaWLo8hOIPzKoERGGU yCL8inpCaov78F6v3F/5vriNXkJDBylTwlsql5bUAk/irzLKMyXgcSgjy3NsU1OUZyU1dnYmIxB0 HdpamGk2IZWnwb+VKCmlTs/V+c7ETSOEOZCRHmsb8ZR5W5lY3Kkqd/aAtK4vJjgocrpHttRWt1Tb ll9ihmKPmWBklDHWFQMJBQZVsdJdmOtYX0yNLceK1o4rX83o5VxePNgpH8wAUvVItY1ikhPt8gpp B7ZReXFut6i37jPZq9lUdvTaMftdwlcylCnBj8xAoNl14AZK1LmKXAVShVQZxEi3sisDnFrcb7EN OSlsgwEcR1rJfHM4enTXrbaJ9S1HyLHmHOKMm13lnz+U0UXOue9R3OyV+9brFIc+X6TBv9kF9A1i oBr8G527j+PHBbqhJai0OQ160ZN2qJB6fpbryVvOA7qhpcs8f43SJtB5YAQKHyz/VigUgjbxNZJN AIZp6AShopWidaIdIrGIXyq6KrovahGZiHuI+wbrmFB2xJOozVyxXQDx6Na0Kki8WWxItQvoskU7 YGgXAFZoRhXhoKKHJFjSLAbJxKX9o6REIpPUSaJJg6M2iDSU3gP7ZMdlF2R3ZCoZItcaHOmGm9I0 nlZAZ2oFPa2gmD1oG24qVMHRJqAztYKeVl54CA5IU6t8XIJfxuvx5ziADqFtRjeJdLCVvp9+gn6J /uf1lz8P4Af6Om3vGVyTCP0YNngfoa9ncK2TUgb6Om1QcgDo67RBmTagr9M2gufCA/o6bVCmDejr tHkJQ4RAX6cNyrQBfZ02SFIF+jptUKYN6Ou0QZk2oGOpQpLqn9MRB/o6bVCmDbTI2nTa4LthoK/T Bt8Ng7aXw0tVUKYN6Ou0QZk2MBpv02mDBA+gr9MGZdqAH30BXafTBmXagL5OG5RpA8vRtahOpw3K tAF9nTYo0wZIujGp06alnbfptEGZNnAUO4fpdNq0tPM2nTYo0wagTttrdhsfGejrtEGZNneo09aD axe8yC6AYhfgQLMLsHcI8ulmF+Aa6G63YELghEC7EMRhHnGVbTawCzC0C1hrFxYVqBuVCJtnF5jD SgnOCcyaP3+etUNQzJExAQEMuxDD1pNByLmYqIDyBesc5u3gNmMdiftFmBqr4t4NjQpUc1spEdon GwTrxEN5S5Y/CFweOJm3LQL4805zV/MeBMYGbuGJuT/x0vvyevCn2gX0cJg3UqAZk4ieRwH97QIS cR8hHhUYsj6QGxoSNjAhrLdDkLdrYEe7ReuC+9uFjgkMD9xuF9Y5PKBpKV56pDdqn23Hwkn32DfG OAtitohn2YcK6vLXCVXU5e5fH9UaJ4QFrrlNHeIQSIk5yA+IeWtqxietExpjOo1K6zB+C/Esu6Ux otXOIZDmvG3kNr6LhV9aV5P4cWkjN1KxtETa6k2I20bnjevTjeMMpnVeq4iJMRy0siTNLCLwK8oY QXggYnBjI/XOlumGO7aE9axPsAtiqPmtGcEmLFbcYrdJeACjFwt6yop7MF8aW2Z/4RAU/YYu8BAw NIzCEVd+jYmw34VED146fpd2fOGua+5byqh9tlnp3QUzn//mmkt139c47E2a5ehVATm7BdTRcXUb d+xzqtlK27arZus+8fFc5M6W7Pv7CvN/26oSH89OfBjWqzUlWEusu5qAzQ9+toy6LEBjNhseMJLY k6LkKUEHhf7SdeEhPK4Psil84GMeJWQexVh4b7Gk+M2Gm5LIKMpIgU3EPJbQU+heFh449xxrpRCL 2bR2TKClYF+Csd2iqjGBv4WEhlSvfSk9zrCJYZUfpljkvsRY7j/HKKkxD6kxNzF3u3md+GGBUioE dg27Pi//1ezdZbFizeCJ8lZpqWuuQ163gksq4jH3nquY8bUp+tatfNcv17YMK63LsC6pNkzYEvZl yrzvT0YX0SkU/gy/tLnFsF8MvpdEchO4W7jlEghaaJDwI1Qt0kTVQt5+1XjepccPQBGvSRXPs8IF ymH4DtV1HiMU56vg+4LfnnQ++eBG2pt9fLsQz8DlQ+dxkeAxZfVXR8vcZKSmSUCabLesWFYhK3th 1vsjo5utDnl1xnb5XQsegUz5gdr6y3UZP1mTcGZA4plJODPEsIIH+EHVQuYV1Wg8m2mMn2YCHv6A uRuHcGZ7LIv5A86fj/GY32PNTIOOK1KtTaYwFywJXNSHMgvkMOnAkT7DbUzg7JlR9POJWxNdbcEU x30JN9LsFrjaBjn49RnfZ1X1WgeDH2xZrDU/Lt294unVPj4rTo9s8I25r9DBUzrBoQUVHFsIRhnd E+Uaqt23KGOBRGPv+3AryuAFvyxaeF520cXedujNjUMc6zYNcTR5vmuIo2dMITLY6ZYn5YQntdT3 suEF38eetzyR677UbXfn9z95CqsY57MCSrYxPDRUu921OEm1877jHGd/YwRlDMbPyZnYfYJdTqTM axtFOTI9CMkMpkpzFtzYdMHXgvLIS+aJZFGO93/Ye2En95HcMxPAim4+wOUwxK40jLszbqzvq6H9 vBE0kufoXm9/QXTdtyH9um+jYXLHMa5f9xpuU9FpaXPEAYzolyTZ9OqWEK7A6u0VM1gqR4fe/Btp Lqnj+g4r3ccfVmpj11d6pMP5RBvbHRE2/ert37P8tRkMJjkxWnETxtChC/pYIJjteNvWGSybPmAG G6qsQMlopshR5CUCISJ9anb9+9RsaOqV+DY627wi4bxgum1HW6MuOQYyD1MbRofIwxMEBo4I3cxg Q66OtG0lAe+TtnMl4H3StoMsgiRtR5/xBNxk6Z3obcAFLFknjQRpKzesz5PZ9LeexvB1M33SeaS0 u7PNANZug7VK925njgXM74eAya3JUtJ+IqgXWHhY1lduM9AiPsLVtir6zEh5qNzGdubIvmwgXwTH F+gMOL6wSboALFtXBdEVcTa2X8RHrHbxlNfLU13mpffonXTLbZM0pEPVxsuLu22X26sGOlGQ6HMj VaEqgNkWO5Qnlov/fBeK/34XCinL4VN/yI5Gqi/tzNHaT+zYtZf+bCy1aGmYCionDh/4wrwTqpEB GWs3Zz2Ju9qAgvdxVz2Z2TrcVeMvASKW0/SZU/BrozY7P/QYrPY0xa+aiVinDVgFtpXCc0xwm/mI iTPNsV6YA3ZRMAYDbQ3yUWwS0R6DtgbZlV2JhbNBW4N8gz2X/TuBkMekQIiuQV6kaY/1BELecDK4 WoGQOu5TriEvhTuAN54HZvI4xNN7PG8zz4Z3nHeBx7/DU/EQoaUwk9db6JpSgmsZdtMwkmEXIlwu lK4Vbox3FQ4UVghnrcNVuJZhV4trGXZZN5TF6a8SU6ipqvxHCSbpubEWm1pyoyb1SNMw7HpvyhH6 HxK6i24IVcrobbX8TAflhi3V2zJrK/dlDBR1qxDNEgWIcrOity8TrcsAmaIiEVtcLYoSvxJtFn+b mdBHPHFPOHTHeCaOFbfcV3UrUQ0o2JFfnZ9CwwpzE1yO5GypFIdLnogpBpL8rpIvT2aL120ZvKdH qgKe7x1lmySFRcryfBVk2D3nl+TWiiHDzvnn9Ok/VyP5XlLHItZh58Ijexsk8TJjSEDYVdhyUlZW ViSpOMVK3WiRGiNLlvWWj7hdvqsAwU+XjipzvzpeNu2098W1W0CrbDtpYnz1uvxn2W2Z9LvN38v7 qTSic9dVlUbFyvIwhbs8SO572mKv2iQV4ofOPoYjDDGzaxQDFIpcT9sGxY2sad1+U6SrFU5mcg3D 7tXdYQCJpTxRyFvlu1QYXq5KzpZrGHYLnirzW1KUudFNcIRBrmHYCdSV6V9GJT8tT48yUE+omHyB 7y3td4SaOgzXuHeoi6UFaovXgwoHHB4ptUSUSpCsRtSK5x3f9PrpJ9yPXovPmtc0TVV8QkYzUVlS aTbIWOPIR5PPXit/xXejP5tDn6GK4lPgCANPy7B7eLWnpRlk2O0q2GGYrHZNGWv8osNAeiU16zr9 pqbTalaSW3k5dvbW2ignazOVOQhRBNuOtTbrY/aQL6KoBOoXis4JlpW5CRfRy+Vz0ddoTRx63izW syXTojJBBRl2u7pUV9aOj/XcZZmVEmVElVoMMXPqUpk/QaqKUjwXfP+Gn1X7RQkU7ljDrFbGzm75 wuwAIyWqWK5h2O2yLA/fMW0Qozh47GFrgTq9zzSPfrVR6tf8kpfxC/oqrpYuY1pimcwDjITZLXfK D9nWpsf+xFCWrMwfxNhlBkcYfKSdd1vkpsQOLIhV/6bw7NtLVf5Nb6Xd88G58O38d4cnkQw71rCf c1N+YnQHMnsf+5RpPWyVv/FL8i/2+tFs3rDxnraefVuU1aMZ028yHKVmrmnYbqwYSzk7bLJ9jQ+R Pz82ePTw8HTLaQPqiLny83YljBi1cQNR2o0r+GhGTsHX/K2TFRw+J4nFHs6+OF2ZEl7RUlky2f4o 4yijMuq4ZxP/IivLc6yvbRQdMuxwVjlj9azguje4fU4BOcJgogytm83I38+ezDnPFrJK0xWO/hxW 11+vR6pzZ9cNaB6d4+Ny+buNLbXVyS3V5cM5FzlenBDvKLCYs5rzsGqBWqEw4h746gTnrA+3hvN4 dnkda8288mJKWlB+IaMPI8NMAHYHhTfZ3FFEcMuLc7/dFMQv4lbVccO4j/07x4TXspQpSLAnq9x+ nP1Bhh3D1MzUzGy+er76F8UbbuWZoK/5k3hmC3lN3A28vX1LLgzTEJBcV062hRjRWF43oWiVwxqP /E0sZ+E0QfGc+Nyg+NW9hNFJmhGGZP536+JIhl20EHTUspN05CQgFpLsJJKcNEQEjmrZSSQ5qUQE dOwkkpw0c9+/wUBrXRevfTkckSwrlP0sS4ZaK74/dKmQvZJRk/ldXsugiR5wlSfz78ggJGGjXAtJ mPPOAlcDSbgl10ESilSfhyTcwz8HSYAC7n8MSYAC7n8MSXDGPg9JWMb+HCRhC+dzkIQj3M9BEqp5 n4MkNAv/BCShv/hzkISpks9BEsJkn4MkrJF/HpLwk+pzkIRf8N/wVtyMTjqZA2hlvpquczKvpoMH 9Ff0jiikopJO5gBamWehpJP5PbQZBcbML5iDmZOY7kzIo0higm3MH5gnmaSLuQFmjYH+WJuBMVQq A8cw6GL+GIMu5r3Zo9iAdNiBBju72IfZ5Wxwk60TrSI1qwBprwFFq8o41zgNHPCCoxOtIjWrwBpu JpcUrSI1q0CbaBWpWQWgaBX0ZNBZMrDaRKuS4xcl869Tkvm0ZL59ZpxPN1Yy/4AgeW2OIEeQnIhk JvgQl5lBMt8wmb82eU3rTQHJsPs+4VJCsoDVFE9LQ7ZcSLD+JTMuZgc/gJGcaNja13YjEnOTPyhj XWYCRTSiGSPuF9vUmIPIdeuLmwIfUWiqZoTBiP4cg96bV7fOTKsRgDzR29QS0QBRnWhmmkLwWuTe RYSK14n42VOT+WhmQm1GDmTY3UqgAKv+yXz8JzF+U3A+v1z4w8XEsIdremfGeR8QdMxOTlHHJyfv EFQJtiev6byoiu+7z5b11e4I4Hv4KrgRtvWrPRVhW9278UgxFuyS4Oga6pBMASXGsiDgdYxpTLZm QOGGUUyn7YUdso9Q+UPM3bNb7TIFtP3HMo6t41ucLOy6ZLeoMKOIarqnkFZ3CDlYtL/oyeHFuzYb RO1X/LLZMDXDrtSs6LqAsiOnSoAYTCijbnEqN9zxek3PqZJkQeubsM0+xQqEhYAvnmTGRUfnHMmB yu6yAc/6PfwldXMZEv0mLS27TAOqGJ5bfeAMZbss/zS1uCIOm9pouCjql19SjQttqVm7ppRRbjhW 0N6WOVYYdJT3vI44lU+7YX3Ho8Jb/mvPqsRZ23o1xduv6oJqWGAphZX8r/ZOpN7mw9fkVfwMAb9K YHD0xzhLSZ8TVYn3hD59kBfrBnqJzydQFovddoDejaHHJqpubMrIWZFTI8g9gFCLxcX3qgQ/P6wV t2x+sd94h2BFTsfDxskpOwQemy8mmo4//rLPQzOfguL60ILW+B9fYjOeKcM2PwzbfDMcc09O4F8S SKlmshd2si3Pz13SDI685pUpZO/GDw6cyTy7umZ5v+k52I+ymueFY8tX5Ayecs6OZpjwes2XTQmq mp63rsCBg5OFP1/PEYJDQpWcHKcepAoXfysG01XX6V+rf6XLRJ3RbNHwnsiDrmJf9IFoFdpgtAWl oKxx4ovofTpE+3vYdbbt9mIq0rEgOXEyd3FEIL5kJcM7ZHHEIjwwOmToPG40MFmQZ7AbsTE5et5q u8llw7vJgvTkNRnX+p8w6wd+tPd5YXcm7wqSed8WtHFmoCyiFe7fD3c+NfcIneXzTB12ClGsOuYx Nk1CGrzfkgDo8E6XZeAjLnYaNuXcwSlr5sgGMgaPaBlvdd4+Od5AXTZ6k6z2zeGz39zPeNFyTNVQ iDOe9etOX7dw26UL+ED/tz8fnMJY4pfvyrjG5QjmIgenHPEDcd+UuEo5PatI57GXUJZ1toqpcmJ6 qeaGXh0TxQUpysFOFlibOLY7qY694Lx12ZgXjUGICjY8be1OANnwwHaHSmlcmjrl3OKkCjZnfpey MY3BXZ2fs8d/cwe2QcZf183g8OfOXZw0ZOuiZMa1oSLWTgQsTorZ6rhjdTJPBL3AvuBqvcCCuIXc JO427mAxSHqUcVlfBBuQNytXScP0q9ge3hEegKpHJLjdTgi1cfn+QkyacXnKucHlScKeexlPWKfs DgtlFZAMW2GXr4RnQlW5dtDp1w6iuXMHlzN+qxhUwbj2SIZXIoPLR94A4M74iklyEvwOXQEBtAWM 0NgCNsqTHil/Bmni3WKpqkK8qprsqoR1kzjgptIebof5lhccANmCGMhgC3IW78Rp/e3lpMcncqtE 91q3yBik+QLyW3da/OW+JtXNpLk2WEqHnurQXfsCvUY+nglmMt/Ik+ij0eGqk/TFqIIO7qEGqDGz Ar1F9CWO0enM7Dh0O30vaoEuGNxz5SBXPGRJSCTU/scZjhERixYEBzKjF3CXeNztkyL+RgXI0y9X dVR1TnVb9ejRj8DKrdjMrfgLt+IqU5WhqVuxdHCzFwWqbr+Zi18/w5GO2vfjhP31PHzD6X0ZOCNh uKVAfgjfcLbvzz62RW51owybEuOM1+XEgjaMjSdVYU3pGQwR7gw/wDAMRsASW3uDvqVjTIKR76yi DA/KzaS3riGJ9PovqnaiVCbXBtykd50gZs9FzfwEyCC0HjUrYp9mz/f+xrYvydMHGqL+AN6NcrsN 8uzbc5A3tpLo1aM0Zq6tFvYFfRUTwim1zzyPTQ+unHesG3HDphN3bMlyCenPyJJILksgZNUb6y6j JHsolq4dvMxFtnxlYMChvgF15m7FWECM7PqZ8l1fKOcbuXJHnXiRfbuBPm7CkYcR2IZvr8julbMG L/tt+aDlptce8fDvkMHLwMhksH788knCxamvh9W5GLFB9+lBYCDLpGm5fK3clBvOBhJ5Z1W9/CuO kcoibfqSfiKKYF+/NRzofneKU8W5ywHPOJeE3bgDuROOW4zqPWqaLyIlZSV1LbS97qynUvhp9ufW 9DniwmMe5FuMGrM3lgd4WpMP8bIJJ16sK+i+buGDAl5yCdQMKO9z5P7R3kdNrzVIXpxCkD5HhlS8 LBl5dLQMigcAUj0AiHX9c1aIaJzI+JdpeQx5Fy/XcVOoAAjTUtKP9dGMd+0wTrlR1M8PsSnt97Wb l+EGl3tjkJ779/cHHS6b3tgdGjHxBxP7xLLDhjYjg7P7MO+bOGBKQ9aYcZdlUzADSUSDDQdQKWEZ ab88fX0deVWNUMJc8pHQKQjFOhex2ous6evYJeF2AyUUuV2vaqBs649YxVObKUYWSV2bKWtNLJIG NVPisrskYM2UNSWlVvFezZT1xGxoM2Ubw9YqPraZQlc3CIkfpzrert/dTNmZfe/eUeLHC7a36y80 UywtkuTEerqoG542UzYSCxrRKN2kxHcPGsVK3cCkzb3ibBU/mUY5k2MV70ejODvOsIoPp1G6qxt4 NMrGQmJuE43SQ92wj0ZRXCbmTtAoX1okyWiUwbbEXD2NgqobWmiUIA9iroM1xVbdgFpTduc4364f YU1ZXXq73tWaUlQKrOLnWCtGe9tNqLhdH289KNL7dv1Wa4rwLnVzg6GowQcRNVAPNSCHGqgVxExF A/V2A3K7gfqSmHnZQO3UjHRqpnZt9kG6NlMHNSODmqkYMYM1U72aEa9maigxE9pMTWxGEpupQmJG 2Ezd3YzsbqYeJWaONlMvNCMXmqlyYkbeTH3ajDxtphrRfBAjGrUHDelBo46l5SBjaUF3vyH2BNWP hvzgR6OG05BwGpVHzPBo1E00ZBONuo+Y2UejnqAhJ2hUGTEjo1HraUg9jdpCzLTQqFbWiJU1FbVG fkCtqROtkYnWVFdixtWaOscamWNNXWmN8Fdajyyi94q3Hh9J77XVmjog5GI6P+1w3Y2sUxf5aYCf VrGjqXRT2su6Tk1dm5pswaAmrMkIKSkNbUpsEjahheBoUw9E3vS0aSzCuXwm53gp4CEbC/chissy pB5pQYZcZqHmu3NWlxaVrjRfLIg/dRZgTc9LE9PTvfAzu7ybSvP9jCdv3uVtnTProK9p7i7vTWkg 36/x+i7vl3X5fi/rdvp3agJKv05NO/27NuX7dSW+B8GEJtud/hiRMIhIKCklFsGIidAmpZ8RstM/ sYlIKCFWLyRyhBI/EPXM90skJo4SCULim9iSfD+0cKe/nEg4SiQ8hYv0IJY1QvL95ERCD4Qo5Skx MZZIGEv8wLlMJHAu7/Q/k5PvdyZnp/9xWOzx0p3+PCIHj8ixsZDIsZFY6T4iYR+RoLhM5FAQi8iI BBmREOlB5Ij02OnfQiS0EAlDYI4hxNai5kR9zHf6784hEnYTa19N7JnVxMqLYClFxMRKIsdKIsd8 mLCRWMdGj3wvhHjWp2g/MMzbJgHwWdSoHtmoHqx8hrx+jhM3B+Le0IeFa95GHbZiXal/QUENfm3q 0Ki+F8FToJcLAaA1qq0ch1CejQA3RtNePfcZ3qju0qi2kL9BlM/esojVRSjQO4WjAXBqVHsCq6mN 6gnKZ8Y1Gn0QYE95P79LozqByD+piMhvDvPDChE/z6h/gTSqnxson1lRkQg5egQ3BsB0PwBXX4Nz RBlL5KhUVahZhgWw1U+b+gC0CAC0Ud3iCZwb1b25chRhjyeuGMPejepuNbiBqSGxnf1roBSNsc+w RjX+RE0lEh48tbJpVHeswUFzU3NTw3MfolhKo5ptZ+zclxLoMxF58PRJkwcdx6cZps43UbwegPza FGZnjCCUrsC7W6PaYCU9mfZrk9XJ+heDrCk+y+iN9H6AY0QD4FyicU9iFR5DkKOm3Q2HIGDqQAMw lt4feNO/smxuCv1yPz2FnreTXkB/+GR7/QsXa4oHeKzpvnYZDI+Za8yCeSEATnmHhThzgzWT7oGh IUOGgeeWjWqioq4hUdF4TGQIA+KLjQKjgc9qYmMb1f7E5r1SX2l4bkAlDt6IRnW0O3Hwog8Ru7BR 7eNiTknoCr7rATqgNE5vLwW6dojpyPFgf3oP8KTJ9eETuAeIXUCtUVtZPHxiVf+if40a6MWfP2cO LCWKNTz87pzJGgE6wHMGaLaF+ru1UWrwB3daW5tanxOLHo5RSFCvw5oaO04bQjkwAnwzGrQ2ORL1 Q8wAx4ZYA3GevJG/EXSuwXWqxcpnrFbimCxVoOaHQQfNGeirVaJRQ7OJe/RQykUPOWrIWQpGxVAM 4clZQ6oTmb4388n1uxDrn0Gs/zZcf6ePrd/Gkzj1HiylGTKIM5lYH3EqcGdsArFJxD4lVrrqqXqW 4jUCiJPP1NCA2C8mjerQp2q7GhwOU3Tzp79svdfSzTiBup7+4Cnef9jCF9vTpHcxVn9vwVdZPyI7 6JVi2pitZSuJ/XJs2o9ytOG5lfKZNVGUmC1HxzWqvYk6DQJdG9WrcbChq4Gp5fUDONhqZd3NEnGX N6LZjeqljeoroO+pRvUiLzlq/lRte7BR7QS+aFSfdjUbXv1Uvc5wknmPRvWKvrzuoISo/Y6n6hYn QFwlqUO7g+uzk8FUtsWXjep53n5ydFcy6GBkQVxionSOHM1PBt0s+71qVJ/MMVtCXI4eq40sZpn3 ZitQyVGzjd0sB1DNz5idPWq2xXKY9w5i8upRs5yJFngn8xN+ioe0AEDcGmbg7gr0AHGmsola+YIJ jWri0nz9qznDlG2Bic2Jy5qzmTgrphhZEFc5J9pHgT53YEw2ssiYaGL72mxMsbXVNNtnkThw/IrB NQMs0B7/abTiABjoNRm6gJewIlH05CU7zOzHNGMwqN+hW/ZEWjrxMdL+zgJwWQCmEh/iTAL+xMeQ +MwD5OUfoc27GpCN0hrt9z1ioZ7EN3HTByHDRowaMWqYZh7Rlv2vTuu+XYnPSBcAcCI6a9NT9LaJ p/3oQkAnPwyXmVNg7sEuZL3hckQyWKTLaEBuEIv44b17JYy7BoC/nwp8gRvwIj5sIskZRIJAsBws AEtAqDbbDFgX50/Xha/9MMBF/VTsY9PE5mm+4X7uT9QtBdaPRUzrrQfR7F0y/ePB12bma5fhw0kg lUqBtEYK1Of3AvVrHOzJzibSYQ5c848onPyCf3hbXdqjPdrjfyMs26M92qM92qM92qM92qM92qM9 2qM92qM92uN/ObSDPMTf/9GRHrjl5Kc92uMfiI40bfyz1TD5VPkfHwv+68eGP1E++Ni0tui/pgIm mvhXygcf+/HvKx+8P/mfFU58TA3g1L9UfmcqEZp24z8tv6MpzZLWoUOHT5dP3p7hIW876qAz9a85 BUw6djDpSDMx+YPy9aZ1l0GnzvAdb1vSf1B+BxPiDmBiamLyiQwf7n/N7u9M7IF3Sf9J+TD+6NYD LNuuft3xB8TBJw/AX3AQjDTxB+V/ME180en6Jf832+6PX//aY/4XXYV/tPzH73+/r89/r/yP3//f n/+vlv//1/Lavvaf2eOfWQtou/1+Oqs23385/rMC/q1b6HuL/PGu+s+2/6Mvw7VX/3vX35/J+Acl fPLXj0x/9D7wr3RU9G75upI/qIG2RfrD/sdnyv9UR0W/yfnEUSRK/1T58P/3+x+fKP/DjJYf9LbA +9n1s3xYqw/7H78rX3ckP5WxLfTL19bx46cAAG33g3erfdf/+Nz2/y4jsVkdOhjobT+5pH4dO2ry fGL7dRV51//4/Tnf9vWxjJaAZmlgqe1tvluorY7aLDTLDvqb8vv+B5Hyu/7HhxX9REbivOtoqd/b I3c/WUcyS8cOtA4dO1j+wfH/sP+hmwS6kYRPZTQh9q4BzUTv+tOrI5mF2PU0IoepyR9ef+/3P/50 Rs0qafoPG/AHbR1BWxaihh1MaZ9a7cf7HR/N+LurH55aHYn1dnwvq/41Q2Yhfjd6t4qP3lY/vGd+ NOMHN9cPe5sf3KKNPtMj/cvj72ii/z+Kjx7nfzPjn17Xh0v+a9n/8nX96WX+RMZ/q/y/Mv6N8v98 Rf+bK/8Txf+j1zZ5Z/3HavDRhuJvLf5j3c+/sXxd9/MfqQCw/OvGyf698vXGybQpf8cV31b+78bJ /ubQGyd7dx7+re/awLvu5z9zAbSNdPxzN4DfTfzdFdA+1/xDxf/ThbdHe7RHe7RHe7RHe7RHe7RH e7RHe7TH/0RAbZQu2kE7M73p9vi/F4bAGXBBMIgBi0EIWAKigZdmejEIBJFgJXAj0uYTOSI1KdFg ATG9RG/5MYCi0dvRxZ8rE4A85LPZ/nT8q+X/1fG/XD7UUYLaTVB6Luqd+hGRbkZOGYN3snTgpDYv jDHgPX0lPtSFgtpOWYDUUMoDpGzSYe08/HQjPhxfdzdnjuMMbxfPWY7eLrMcZ7p4AW1es3e/u3u6 +Lq5+HkBFkWXDpNnuM2a7uTo5QLadKTgYKgJ8e2oXYdum3SnV0ezf3WP/N+K/we0yiLpAAAAAAAA AAAAAAAAAAAAgQAAAIIAAACDAAAAhAAAAIUAAACGAAAAhwAAAIgAAACJAAAAigAAAIsAAACMAAAA jQAAAI4AAACPAAAAkAAAAJEAAACSAAAAkwAAAJQAAACVAAAAlgAAAJcAAACYAAAAmQAAAJoAAACb AAAAnAAAAJ0AAACeAAAAnwAAAKAAAAChAAAAogAAAKMAAACkAAAApQAAAKYAAACnAAAAqAAAAKkA AACqAAAAqwAAAKwAAACtAAAArgAAAK8AAACwAAAAsQAAALIAAACzAAAAtAAAALUAAAC2AAAAtwAA ALgAAAC5AAAAugAAALsAAAC8AAAAvQAAAL4AAAC/AAAAwAAAAMEAAADCAAAAwwAAAMQAAADFAAAA xgAAAMcAAADIAAAAyQAAAMoAAADLAAAAzAAAAM0AAADOAAAAzwAAANAAAADRAAAA0gAAANMAAADU AAAA1QAAANYAAADXAAAA2AAAAP7///////////////////////////////////////////////// //////////////////////////////////////////////////////////////////////////// //////////////////////////////////////////////////////////////////////////// //////////8AEgQAeJzsXQl4E0UbniSb9G56U9pCWjkEhKbcl5IC5RQMFoqCaCgtlyAptJwFUrQC YjEgBVEr8SooWgNIAZUaRKig2IAiXtBwF6iScrbKkf+bHGWbbrKbNsXyP/MxL9nszs688858M7M7 6e5hfeDJ97ZGnEJ21gcJ0F2TFxLR9gkBPNuXAIT41u93TSaTbTdljbcYDpwCmIg1arsDWMSz1KEH VOgd6z4PqENPgBfAG+AD8AX4AfwBYksTQIGAIEAwIAQQCggDNAGEA5oCIgCRgChAM0BzgAQQDYgB PARoAWgJaAVoDXgY0AbQFtAO8AigPaADIBYgBcQBOgI6AToDugC6AroBugN6AHoCegF6Ax4FPGZu 2wjJAPGAvoB+gP6ABMAAwEDAIMBgwBDAUMDjgGGA4YAnAHLACMCTgETASMAoQBJgNOApwNOAMYCx gGcA4wDPAp4DKADjAcmACYAUQCpgImASYDJgCmAq4HnANMB0wAuAGQAlIA0wEzALkA7IAMwGzAHM BcwDzAcsAGQCFgIWYd8EqABZZu5K+JcBdTEA0s2AtPAZ3C0MvN3WlrDvB0fwzfv3WA4PpMeN7rBp +Vdt9vME+IvAso8HSk43l7pu5o34PHp75npeiPVTAOUeASoMgVodVYf8A6AX5CFLcbjmj/vUjL8t OvGg5SmhVtNAhwnmmnbNgupQfsx3pnUb+z3f+mkD5sfk/3cbtisi9h8YtAW+p137mMcL5lzvPDgq 8La0fXvfx/3/8Kkps5TpykkZ0aOnpk9VRifMSp47dcZk87gxesjIIfLobrFx0SOnJKdNTDcPKOZo sdZosd3Q9Z7bZtLJ2eWPPc+vHrOMuyY8Z+Ez+kl9zsd7+L/8+EtebGTAmjc80SPt//kMjy9Cu31/ wb6PrP6G/TLecnoWHnvwOIvHGDzWjrOmiccEPA7jvh6PxbnW/WcpyxjsRVnGV9s8zdG2r3X+tgTw IuAlQDbgZcBSwDLAcsArgBWAVwE5gJWA1wBqwCrAasDrgDVWLmsB6wBvANYD3gS8BXgbkAd4B7AB oAG8C3gP8D7gA8CHgHzARsAmqy4fAzYDPgF8CigAfAbQArYAtgK2AT4HbAcUAnYAdgJ2Ab4AfAn4 CrAbUAT4GqBDlnb6DWAv4FvAPsB+QDHgO8ABwEHA94AfAIcAPwJKAHrAYcARwE+AnwFHAb8AjgF+ BfwG+B3wB+BPwHHACUApwAA4CcAT79OAM7gOAecA5wFlgAuAi4BLgHIAbit/Ay4DjIAKwBXAVcA1 wHXADcBNQCWgCvAP4F/ALcBtwB3AXYB5oIDGwMMdAEAAoABCgAjgAfAEeAG8AT4AX4AfwB8gBgQA AgFBgGBACCAUEAZoAggHNAVEACIBUYBmgOYACSAaEAN4CNAC0BLQCtAa8DCgDaAtoB3gEUB7QAdA LEAKiAN0BHQCdAZ0AXQFdAN0B/QA9AT0AvQGPAp4DNAHIAPEA/oC+gH6AxIAAwADAYMAgwFDAEMB jwOGAYYDngDIASMATwISASMBowBJgNGApwBPA8YAxgKeAYwDPAt4DqAAjAckAyYAUgCpgImASYDJ gCmAqYDnAdMA0wEvAGYAlIA0wEzALEA6IAMwGzAHMJeH+3GY/wEWADIBC3mWvlxknedn8Sz9AfZr 2/y+bQy0c7xhtUo4mAgT6egBwwfi2Hje0k1gmRfgfmm6Nd4euAjYE27xE9xO6LanqQDtaULBvHgI zJOHmOfOCTDTTIZZ6lSYdU62xmtl5QDjUZZtG64ls/D8Os3KNc9Uc46z6l5e0X0hzamQ6nTU2Gwv 414/lrP2fPSnub/H9tGQE2ZcrwxCSZKqueGtqubi8YeqCkKdQvFU35LHeMB2iIePHYXzPXZ0GY8/ sUUDrsM2TvdoYZfxON4e6zFsqXPMU77qeuDRtmEsympD24/nBgE0rl7W78Oh/TzhZ7mGi0L3xhv6 fQR6+kEM23hcwtd/4dZtbLZqb0WLg6+z8DUgbrvPwbj3NOz8RGS5/yBGm0Sj+BgTKDHC2CSity96 WehtrQ0t/abWbZy3rQ3i/Uxt0FpON7TBE4i6GoSWi/ai3VctdR3fzlLXh+B78+9uzw332IuyHj+B ctV/mj/Ntqn3eHwN8nEImofj4k/7lKtOHTdrig2fi2HLw9aecC3Y2pOAFteWJg/iFxb+iXrAJ06P umY5vw3tfJhoVLdHe5627S7Xgszfmwa1GI81Pgzfny9C8+zbINabzf/dp/1xTtrjcuNy4E+zeVq0 qqBOoHiIlyWFdnndcv5h6/m4zFj7KojrDu1xemza2/O0bXPVnu4b4NJZYdbtcLv6CbduY6P7M76e xo0Q3/vwAefC95cWCC3HkpKWxS8ZuyweH5sktMx/h9nVhm2Ojee+tqvkZBgv8B0KPD/G93dGWj9t 6IRq9iP2fPA9EHwP5yXIMxQiXLXjg4+VCi3Xvs74JACfFDQbOE003zcZTMvfnleUHSe6prZ+0lbf TP2kALH3k9Q9jjXOxfvp/eRpD0s/KfW29ZOPeI/iY5z3ECOMR7yZ+kkBYu4ncfpM/STe3xj6ySnW /meKrZ9E97R2ZDZfxXa/fNWep22bq69ivW3aYz6NQftaY5S1nwz45bjDfnKKm/vJKRy0r+8YRfcN R/0khbj1k2pPSz9517N2P3nJ8/72k/i+tYeXhc8cr5p88LEJXlz6yWToHZPRBEA69JU2PlEMfHog x/0kvU+KsXya72dgTX+DCpCDwN2h8/xKZFmDsNULvW9iShvR0qbnAzpn0esrjLaN7/mnzryXb1fI 8wvRve82S0aW+/r4Hk4bax3/lWY5Fo0OoeeAyDS+5ZgtLs/a8+B4+/wtsNeVmEMjGrEb0YjdiEbs RjRiN6IRuxGN2I1oxG5EI3YjGrEb0YjdiEbsRjRiN6IRuxGN2I1oxG5EI3YjGrEb0YjdGkojxGYN lG9DGNGI3dzP1aYBU8jOxsFk/XWgI8kam75EIw4lcltC1WYRgh4Y9bILNTRwJXLDS0Y04lA0NyTB sWAmJvkY1WSNZheYlHKjiEQjDgWs5/kmLhq5Wua6BXsPtPdNotEDoZEbfaeBRSQaNWaN6ulrToR2 XNr6iEU0amQa1VCKUSxX3YFz8ZzFpxvR6AHUqDZdV8VybCxZO4hpT4xo9B9rVO0/TuPQmdbD9Zzq UgfjIjfziS5mQzRiTvqeKFw0YqDlyPVcmRZwLxJHVi7E55Ai0YhrunaBxZyQq6MDulsp7kY0Yrc6 9tnW4OT6w7FezhywUSpFNOJQatdi0wpurxRjl8zkl8wJuu53900sohGHUrsWu1Z5HSrhSK/qjpsx ZY5+RzT6P9OotlJMcRjdsGYhmVPiJpa7NalVatdiE43YYxONnMWzK0PtjsPukBMx6fEdqVk7QaLR g62RMy9gDYzGOOQ90L5GNKqrRqy+xiqZXflrK+VEGqLRA6sRF8a1e1wnbuXEubjnSDR6MDVCTAVg LRIXN3RJo2qF74sRjditwXytbm5INPq/1cgaz+1KMQrHTZ0GVYSh7FzjEY24xrYzN2rkJEGaNZAK LKV2LTbRyOWTnVit4rEo69jcVdS6F7NeJxONuCZhK0qNwLiTfohJmvqTaQgjGrGb+zWqcd1K22N/ SVtTo/rTaDgjGrEb0YjdiEbs1lj67PrTaDgjGrGb+zTibkxTx/rTaDgjGrGb28g5k6DR3OWomxGN 2M39LOleVXvPA+FcdkY0YrcHiet/ZUQjdiMasRvRiN2IRuxGNGI3ohG7EY3YjWjEbkQjdiMasRvR iN2IRuxGNGI3ohG7oYdQ9XsbTfR3P9relyu03u9hel8uPkZ/n+VTAAo21AKEHqcQCoGTDgN6wf4A 0WtCsWia0Es0WGgSthZeFYqEp4Rl1I/C76kdQi2VJ1xPZQlfoiYIZ1J9hSlUtHAMdYdKpP6g5NRO agS1lkqi0qlnYc/zVCdqARVKraT+FXxAnRLspg4JfqO+ENykNgvChBpBd+EbgjHCNYKFAA1s74V9 Z4WbBSLRF4JWokOC/qJTAoXoX8E8USilFnWi3hclUttE6dTXorXUd6Kd1A+iP6hDojvUQVG08FtR X+Eu0QThZlGW8E1RnjBbtEM4XfSjMFF0SthVdFUYJjIJb0CZfoaybRXiMjK901dYU7M6a9sSkMJH aC4cfA40XiK0vGNzMj9TOFawiFPeTO80FqKa1orhXJwmfr9xe8CTkPc4OLgGPjcIre/EFYwDvRcJ nxRsEOJt+nlM7zPGadPfGersvaOlkM8eD4SGglDtve/fe0dt+Q6GfNt53/tuM7b3jk4j7x11txGN 2I1oxG5EI3YjGrEb0YjdiEbsRjRiN6IRuxGN2I1oxG5EI3YjGrEb0YjdiEbsRjRiN6IRuxGN2I1o xG4NpRFiswbKtyGMaMRuDfb7bcYHjWTXemQEk2SNTV+iEYcSuS2haqv9pyOsD0Syk8WlyA0vGdGI Q9HckATHgpk4PL+EYzQOf+LkRhGJRhwKWM/zTVw0crXMdQv2Hmjvm0SjB0IjN/pOA4tINGrMGtXT 15wI7bi09RGLaNTINKqhFKNYrroD5+I5i083otEDqFFtuq6K5dhYsnYQ054Y0eg/1qjaf5zGoTOt h+s51aUOxkVu5hNdzIZoxJz0PVG4aMRAy5HruTIt4F4kjqxciM8hRaIR13TtAos5IVdHB3S3UtyN aMRudeyzrcHJ9YdjvZw5YKNUimjEodSuxaYV3F4pxi6ZyS+ZE3Td7+6bWEQjDqV2LXat8jpUwpFe 1R03Y8oc/Y5o9H+mUW2lmOIwumHNQjKnxE0sd2tSq9SuxSYasccmGjmLZ1eG2h2H3SEnYtLjO1Kz doJEowdbI2dewBoYjXHIe6B9jWhUV41YfY1VMrvy11bKiTREowdWIy6Ma/e4TtzKiXNxz5Fo9GBq hJgKwFokLm7okkbVCt8XIxqxW4P5Wt3ckGj0f6uRNZ7blWIUjps6DaoIQ9m5xiMacY1tZ27UyEmC NGsgFVhK7VpsopHLJzuxWsVjUdaxuauodS9mvU4mGnFNwlaUGoFxJ/0QkzT1J9MQRjRiN/drVOO6 lbbH/pK2pkb1p9FwRjRiN6IRuxGN2K2x9Nn1p9FwRjRiN/dpxN2Ypo71p9FwRjRiN7eRcyZBo7nL UTcjGrGb+1nSvar2ngfCueyMaMRuDxLX/8qIRuxGNGI3ohG7EY3YjWjEbkQjdiMasRvRiN2IRuxG NGI3ohG7EY3YjWjEbkQjdmuQ946+L0BotwdCD8NJ/bwt7x1tITJ5xoiOe0aJvvAMEq33FIrmel4X Pu15Utjb86AwxlMr9PJcI7zukSE84zFK+ItHZ+H3HgHCbz3+pvZ4fE9947GJKvZ4mdJ7pFJ/egyg yj3aUHc9/KhQz+uCWM9SwRDPQ4LJnkWCZZ7bBB95bhb84LlR8Dcg0GuzoIvXNsForyLBXK9Dgje8 SgWFXtcFh738qEtebSjkPYAK9k6lWni/TEm9N1E9vL+nHvP+m+rjHSDs7d1Z2MV7lLCdd4Yw0nuN 0NtbK7zhdVB4wuuk8Fuv68KNXkLRMq8g0WSvKNEArxjRQ164jFze/ensHZvf+SBznXT2R+g1/v17 x6Yt346Qbw7/3nebkXds3ncjGrEb0YjdiEbsRjRiN6IRuxGN2I1oxG5EI3YjGrEb0YjdiEbsRjRi N6IRuxGN2I1oxG5EI3YjGrEb0YjdGkojxt+V0q2B8m0IIxqxW4P9VpnxoRrZtR6PwCRZY9OXaMSh RG5LqNpq/5kE68N/7GRxKXLDS0Y04lA0NyTBsWAmDs/q4BiNw5/zuFFEohGHAtbzfBMXjVwtc92C vQfa+ybR6IHQyI2+08AiEo0as0b19DUnQjsubX3EIho1Mo1qKMUolqvuwLl4zuLTjWj0AGpUm66r Yjk2lqwdxLQnRjT6jzWq9h+ncehM6+F6TnWpg3GRm/lEF7MhGjEnfU8ULhox0HLkeq5MC7gXiSMr F+JzSJFoxDVdu8BiTsjV0QHdrRR3IxqxWx37bGtwcv3hWC9nDtgolSIacSi1a7FpBbdXirFLZvJL 5gRd97v7JhbRiEOpXYtdq7wOlXCkV3XHzZgyR78jGv2faVRbKaY4jG5Ys5DMKXETy92a1Cq1a7GJ RuyxiUbO4tmVoXbHYXfIiZj0+I7UrJ0g0ejB1siZF7AGRmMc8h5oXyMa1VUjVl9jlcyu/LWVciIN 0eiB1YgL49o9rhO3cuJc3HMkGj2YGiGmArAWiYsbuqRRtcL3xYhG7NZgvlY3NyQa/d9qZI3ndqUY heOmToMqwlB2rvGIRlxj25kbNXKSIM0aSAWWUrsWm2jk8slOrFbxWJR1bO4qat2LWa+TiUZck7AV pUZg3Ek/xCRN/ck0hBGN2M39GtW4bqXtsb+kralR/Wk0nBGN2I1oxG5EI3ZrLH12/Wk0nBGN2M19 GnE3pqlj/Wk0nBGN2M1t5JxJ0GjuctTNiEbs5n6WdK+qveeBcC47Ixqx24PE9b8yohG7EY3YjWjE bkQjdiMasRvRiN2IRuxGNGI3ohG7EY3YjWjEbkQjdiMasRvRiN0a5B2b+3ws+4b4I7Sab3nH5gD/ J3j9/R/hPebvyevsX4Za+xejUP9NiOe/Al3yS0OH/Z5G2/36o3V+ndAcv5ZolF9T1MkvCIn9AlC5 bwA66BuKNvk2R0t926FJvr3QEN/hKNY3FYX6LoLM3kR/++xApT7H0M8+legHnwjeAZ/HAON5P/i8 xPvZJ59X6lPM+9vnPA/5UvxQ32h+rG9P/hBfOX+S7wT+Ut9Z/E2+i/kHfZfxy31z+GK/lfxOfiv4 o/xe4s/xm89f5zeNv91vDP+wXwL/kl97Ps8/iB/qf4PX2v8YlGUHlCkXyjaTh8vo6js2E+EzjWeJ k2eqebNtFa96M7ovmoWmomQ0nfFuHBc7gairuLb2ooDYEygccAi+Z3x3ey7eN+3A7bmeor1oL3zP Vf9pPm4zAUvKUx4/YX2bJkL4XIz1vL1mYAsw/78XdQrFr+bciyhkaS84Hm53mA8PuNy5+ifqAZ84 PepaEEqSVM1t06pqrrmh4pbW3HI+NsyPztO23eVakPl706AW47HGh+H780VoXuoc86tMa7Rd+vtI be8dDbeLQ383Kb2tjwKsAvyKxfFA6AeIrPSwHEtKWha/ZOyyeHzsGdgH7oCG2Wlmc8BUc1qzoGZn oHSo3WSUgZTwPR32PwbA72DtYUUnZHk/Kt6Osn7GWPfTuTG9Z5Xpfa5ecEAOlZEBfvqV6P69z9WW 7yzI9wvRve82Y3uf6/Pkfa7uNqIRuxGN2I1oxG5EI3YjGrEb0YjdiEbsRjRiN6IRuxGN2I1oxG5E I3YjGrEb0YjdiEbsRjRiN6IRuxGN2K2hNEJs1kD5NoQRjditwX4Xz/gAl+xaj+Jgkqyx6Us04lAi tyVUbbX/JIf1QVN2srgUueElIxpxKJobkuBYMBOH58JwjMbhT8fcKCLRiEMB63m+iYtGrpa5bsHe A+19k2j0QGjkRt9pYBGJRo1Zo3r6mhOhHZe2PmIRjRqZRjWUYhTLVXfgXDxn8elGNHoANapN11Wx HBtL1g5i2hMjGv3HGlX7j9M4dKb1cD2nutTBuMjNfKKL2RCNmJO+JwoXjRhoOXI9V6YF3IvEkZUL 8TmkSDTimq5dYDEn5OrogO5WirsRjditjn22NTi5/nCslzMHbJRKEY04lNq12LSC2yvF2CUz+SVz gq773X0Ti2jEodSuxa5VXodKONKruuNmTJmj3xGN/s80qq0UUxxGN6xZSOaUuInlbk1qldq12EQj 9thEI2fx7MpQu+OwO+RETHp8R2rWTpBo9GBr5MwLWAOjMQ55D7SvEY3qqhGrr7FKZlf+2ko5kYZo 9MBqxIVx7R7XiVs5cS7uORKNHkyNEFMBWIvExQ1d0qha4ftiRCN2azBfq5sbEo3+bzWyxnO7UozC cVOnQRVhKDvXeEQjrrHtzI0aOUmQZg2kAkupXYtNNHL5ZCdWq3gsyjo2dxW17sWs18lEI65J2IpS IzDupB9ikqb+ZBrCiEbs5n6Naly30vbYX9LW1Kj+NBrOiEbsRjRiN6IRuzWWPrv+NBrOiEbs5j6N uBvT1LH+NBrOiEbs5jZyziRoNHc56mZEI3ZzP0u6V9Xe80A4l50RjdjtQeL6XxnRiN2IRuxGNGI3 ohG7EY3YjWjEbkQjdiMasRvRiN2IRuxGNGI3ohG7EY3YjWjEbqzvc7W9x5Xpfa54n6P3uT5OWd7n elhke5/ra8L+/tOEj/kPFnb2by1s7S8ShvqXUTz/76lLflrqsN96arvfS9Q6v5nUHL8UapTfGKqT XyIl9pNT5b4jqIO+SdQm32eppb7PU5N8F1BDfFdSsb4fUKG+uynk+xv1t89NqtQnTPizT3fhDz5j hAd8FgI0sL0X9p0VlvqIRH/7tBIh3/6iUF+FKNZ3nmiIr1o0yfd90VLfbaJNvl+LDvp+Jyr3/UEk 9jsk6uR3UDTK71vRHL9donV+m0Xb/d4UHfbLFl3ymy7i+SeKQv27ilr7h4k6+9+AMv0MZdsqxGVk ep+rCNV8n2tdtW2Jv8OBuZCIH2SgElresRkgyhR6eWZyypv+bs6HkeW9nvT87TnQ0xwNn+0xH/gy FhLrD5958Gl+P6vnWGF/4MDzzBPibfp59PzpvOjvDK2PJhI4cAsyivRASOBlKVsLkcmziYfJs6Hr I8rDUh8taXlHQ77RHtzqoz55z4bMKMgzTYCQvzXv+XxPr+kCb6/71RZSIe+LsGMjfD7iZWkLqYKL nhuBQ6rgES+8fT/aAlNZPKznmKz3sGnvnEWYfxxABqSeA9Iab4Q+EVne4dvfc5PobW+MCVR/T4xN Ii56tqGlTy+P7Z3NeH+BqeYN9YZ4Z/Ny0V60+6rlfcjx7SzvQ8bvbm7+3e254R570fqhlnch40+b 1eWdzbY8wmnvXLa9s1lAi4uPfRyC5uF3Np+IPM75nc32PG3bXN/ZjPW2aY/rqjFoj8uNy4E/zeZp 0arNheMoHuJlSaFdXrecf9h6Pi4z1r4K4rpDe5wem/b2PG3bXLWn+4YHqvmOaZufYF9leo+2CLG/ R3sl5HYZ/HYE+O1QYc33aONjPYXs79EeAfWN36Q92fx/GpqC0s19WndALLL0QbEM2xjtAE1QzX7H nid+D3UZcEkELjehb/zMjic+9o7Qoo8jnuPMaSWjCdAyJ1r5dabxCHKBz0LA75DnDjxmg25Bopp8 8DGeyPLOb2d8RoCnpKAMNBt0w5y6W/OycaDr1c6OE70d2Pp5vjUPpn7evhy2se8N4D8HyvE6VHKW dR70tvcC4Wsw72Pqq+l5Oxr7+KimtWI4lz72vQx5PwOJHfCx1CMe+5b6PGOee77s844Qb7ONfTht +lhBj2/ru3CcvPvQd2HfD4iFvgaA+6wM6LPwvmkHbs/1hH5tL3zH/QA+brO6jBvreXvNwBZg/n9v dd9lnkuie30X5oP7rjtX/+Q8bmB+dJ62ba59F9bbVj88VLPNMvVXfOS8v1oOqMSdETh6MFwTKT1q +h0+9gzsgybt0O/Gw3YCUoLf9YP/lWgaeFcamo8Gw/4egE6A/tZt2/co2rYjH6S3txjLJ0hpKc+V Zgi1hQp5BA7EA79Qa1xc5/Q2y5Q2oqVNzweSyaJrFUbbfhJrFX4v37aQbx+Pe99thvsV7H8jrPWE +9kAayOLRodQTyAYJrAcs8XlWVshjrfP3wJ7jf97I1yYjXBhNsKF2QgXZnvAuBTl5+akj01PGTot sWfK0I71Ce7gYjKWlpYUlxYXlhYVlBbllxXmlRXm2kIOBGNBtjXkZ1pCZX46DnnTKvNSLMFtXMpo obRGKKsOx2qE0hJ6cBuXvJpcHPByyuh+cKlFqjYjK5dp7uJSabQPRmNtRscKcnDIz64O1XTcxaWA iUsNXk40sqoD7ditXMqKC8qK88uKcCgtyisttIRcSzhWkGvVxapOjlkXKyN3cql0EJxrdK8FgZe7 g0sxcDHhnM26FDhQJ49RnWo67uVSOziXqaY67uZiLCksKy50SZ1qOtAl14lLQUrHgrEQ2kLIT4yB kDc0HEIuhH4BOf08s3uaQ0eUaQ5MStmrU1cuRdP6GQtyLaGsIKcsH4fS/OzSvOxjeZkQSnLTLaE4 Nx243FOnyF6d6p4Ixqw6cynMtW8IxupQVmkJZWVAJ7tjdQp26tToFevOZaixsLYf0+kYLXQwl56e 9m2Hpk41IxjR68SlON3MxVRZkBiTM7RtcVGBhQ58TewZUAhN2UYHuOT09Kx5NrM6MM2oM5eifLyR ObY0vR/QMYIK4EYpPWEb6MBsxszFCG0HmrLlrJL87Nqhmk5duZRkJlaauZRkTyvLTbfSOVZSmplo 2c7LSbdIY+YSYNWjpADqyFiM/zfPsfJqcCmqG5djmWMri/BSwLGc9MLMlGN52RYKxdOGArvClJ7V 0gCX3KH3uJSC0wGLoryaXCrrwyV7bGUx5lKamwlcyooKgALQye4XUwadmJlXfk4mcAFHh36nmoux 2KqLmZQdl/w6cSnNTqksLsQb+TnABXwXNoBFztCYMpj1FhVYpIEWhLkkxlRzqSw7BlVJG7nAre6F unHJwVxAlDLc1+UWZKaYxZqWO7QtcIE9xeaagrqDPjCfxsXsQSagA0M7XDpYAswVqq8hXOdSljOt sqQIuEB7gVCSnlhqbj55iT1xlumJJZljIUDzqcUFj1qgDnAxHisylhSVlZjHr6ICViIOuBhzMRfH 59zrAGGEyB/btppLdcDDqXlIrCw9VnmsmAsRh1zSnXK5RwcGrAIbF7rRuXAk4oxLMdCxhuJCcygA R4d+B7pB6JXNIdfMhSGr4sJ8C5dpuFq5mgMutjAN2k5ZTgq0ZnAu8PVjuKUkQscMAYZQCDDBqJ1C iZkLNFzuRBxwsbNjRYUuJWnhgvtmF40bF1ooKHEUCi0h3xJcJcKNy30zwoXZGhcXtt/E2ZbemNYr 6b9Foa9X9vFHKFWIUE8xQjOs65X9/KcIu4qfZ1yvpK87OVqv5KGa1orhXPp6ZQfIewgkpoTPV63r lR3EQ4RK4NBB/KoQb7OtV/KQ4/VK+/WjgAiEJkAmmSBaiej+rB+lzryX73zI95Do3nebJQNSUM31 o7/SLMfw+tHzQGQa33LMFte2foTjNd71o0ZrRCN2IxqxG9GI3YhG7EY0YjeiEbsRjdiNaMRuRCN2 IxqxG9GI3YhG7EY0YjeiEbsRjdiNaMRuRCN2IxqxW0NphNisgfJtCCMasVuDPXeT8QVR2bVe9cMk WWPTl2jEoURuS6jaaj/yl1Ev5qfeO9bXUeSGl4xoxKFobkiCY8FMHN47xTEah0dTu1FEohGHAtbz fBMXjVwtc92CvQfa+ybR6IHQyI2+08AiEo0as0b19DUnQjsubX3EIho1Mo1qKMUolqvuwLl4zuLT jWj0AGpUm66rYjk2lqwdxLQnRjT6jzWq9h+ncehM6+F6TnWpg3GRm/lEF7MhGjEnfU8ULhox0HLk eq5MC7gXiSMrF+JzSJFoxDVdu8BiTsjV0QHdrRR3IxqxWx37bGtwcv3hWC9nDtgolSIacSi1a7Fp BbdXirFLZvJL5gRd97v7JhbRiEOpXYtdq7wOlXCkV3XHzZgyR78jGv2faVRbKaY4jG5Ys5DMKXET y92a1Cq1a7GJRuyxiUbO4tmVoXbHYXfIiZj0+I7UrJ0g0ejB1siZF7AGRmMc8h5oXyMa1VUjVl9j lcyu/LWVciIN0eiB1YgL49o9rhO3cuJc3HMkGj2YGiGmArAWiYsbuqRRtcL3xYhG7NZgvlY3NyQa /d9qZI3ndqUYheOmToMqwlB2rvGIRlxj25kbNXKSIM0aSAWWUrsWm2jk8slOrFbxWJR1bO4qat2L Wa+TiUZck7AVpUZg3Ek/xCRN/ck0hBGN2M39GtW4bqXtsb+kralR/Wk0nBGN2I1oxG5EI3ZrLH12 /Wk0nBGN2M19GnE3pqlj/Wk0nBGN2M1t5JxJ0GjuctTNiEbs5n6WdK+qveeBcC47Ixqx24PE9b8y ohG7EY3YjWjEbkQjdiMasRvRiN2IRuxGNGI3ohG7EY3YjWjEbkQjdiMasRvRiN3c/m5MCjYOwc4e FEJPhiL0NZzYC/YPD80UPh76tHBAaFdh79BgYWzoNapZ6FHKK3QndTXkHer3kJep3SFp1IYQBbU4 ZAQ1LmQg1StERjUJeZS6HvwYdSS4H6UNfpx6LfgpShk8mXoyeAHVLXgV1Tx4M+UZvJ+6GXSGKgsS CP8Maik8GjRIeDhoKmAlbGth30/CsqBrwptBwSLP4I6i5sHDRN2CJ4meDF4oUgarRa8Fa0Ta4E9E R4K3i64HfyFqEvKlqFfITtG4kC2ixSEbRRtC3hDtDlkq+j1kpuhqyBiRV6hM1Cz0IVFsqEDUO/Q0 lOkbKNsGIS4jl3d/MsXB79BMBKTxLNt5ppo33lbdezlodF80C01FyWg64505LnYCUVdxze1FAbEn UDjgEHzP+O72XLxv2oHbcz1Fe9Fe+J6r/tN83GYClpSnPH7C+mZNhPC5GOt5e83AFmD+fy/qFIpf 07kX4dd48qxxcRvEfHjA5c7VP1EP+MTpUdeCUJKkam6bVlVzzY0Wt7rmlvOxYX50nrbtLteCzN+b BrUYjzU+DN+fL0LzUueYX2taXSe4TLZ6wHxs9YC370c9LAetd1+1lDG+naWMuD6ag/7hHnvNGuAy 4U+zee41c47ffxzFQ7wsKUJx1y3nH7aeX3XquLkeqiCurR5seYTTdLTVg61ebfXwcQiaR68HnB5b PdjztG1zrQeKto3fEWt7F2y4XV3R3xdL739GARYCWgcj9DA42o0QhNYJLceSkpbFLxm7LB4fe1Fo ea/tMLvasHWK42BbjtLQRDQDDYX6TYWtJ2BfDxo62X2n86CXoT7vGNYEIvQCcH0TDs61vmP4g8CZ wrVB6fftHcMrIZFESEwPn7nWdwyvDEoU6oHDyqBcId525zuGPaHSH4VM+sEJIz3uzzuGl4ffyzce MnnS4953m8Va9aC/YzjA2tHhdwz3BIJhAssxW1zbO4ZxvMb7jmHChdkIF2YjXJiNcGG2B4xLUX5u TvrY9JSh0xJ7pgztWJ/gDi4mY2lpSXFpcWFpUUFpUX5ZYV5ZYa4t5EAwFmRbQ36mJVTmp+OQN60y L8US3MaljBZKa4Sy6nCsRigtoQe3ccmrycUBL6eM7geXWqRqM7JymeYuLpVG+2A01mZ0rCAHh/zs 6lBNx11cCpi41ODlRCOrOtCO3cqlrLigrDi/rAiH0qK80kJLyLWEYwW5Vl2s6uSYdbEycieXSgfB uUb3WhB4uTu4FAMXE87ZrEuBA3XyGNWppuNeLrWDc5lqquNuLsaSwrLiQpfUqaYDXXKduBSkdCwY C6EthPzEGAh5Q8Mh5ELoF5DTzzO7pzl0RJnmwKSUvTp15VI0rZ+xINcSygpyyvJxKM3PLs3LPpaX CaEkN90SinPTgcs9dYrs1anuiWDMqjOXwlz7hmCsDmWVllBWBnSyO1anYKdOjV6x7lyGGgtr+zGd jtFCB3Pp6WnfdmjqVDOCEb1OXIrTzVxMlQWJMTlD2xYXFVjowNfEngGF0JRtdIBLTk/PmmczqwPT jDpzKcrHG5ljS9P7AR0jqABulNITtoEOzGbMXIzQdqApW84qyc+uHarp1JVLSWZipZlLSfa0stx0 K51jJaWZiZbtvJx0izRmLgFWPUoKoI6Mxfh/8xwrrwaXorpxOZY5thLqBTZy0gszU47lZVsoFE8b CuwKU3pWSwNccofe41IKTgcsivJqcqmsD5fssZXFmEtpbiZwKSsqAApAJ7tfTBl0YmZe+TmZwAUc Hfqdai7GYqsuZlJ2XPLrxKU0O6WyuBBv5OcAF/Bd2AAWOUNjymDWW1RgkQZaEOaSGFPNpbLsGFQl beQCt7oX6sYlB3MBUcpwX5dbkJliFmta7tC2wAX2FJtrCuoO+sB8GhezB5mADgztcOlgCTBXqL6G cJ1LWc60ypIi4ALtBUJJemKpufnkJfbEWaYnlmSOhQDNpxYXPGqBOsDFeKzIWFJUVmIev4oKWIk4 4GLMxVwcn3OvA4QRIn9s22ou1QEPp+YhsbL0WOWxYi5EHHJJd8rlHh0YsApsXOhG58KRiDMuxUDH GooLzaEAHB36HegGoVc2h1wzF4asigvzLVym4Wrlag642MI0aDtlOSnQmsG5wNeP4ZaSCB0zBBhC IcAEo3YKJWYu0HC5E3HAxc6OFRW6lKSFC+6bXTRuXGihoMRRKLSEfEtwlQg3LvfNCBdma1xc3L0O j9ePhoQilCpEqH8YQjOs60fDQ6cIZWHP37f1o+6Q9xBIbD58vmpdP+oeNkQ4Hzh0D3tViLfrs35k W6vFcR7UNXPbWi02d6yZc1mrre+aOc7fVj9CxL5uy0PO121zAYPEyLxYdygQIaVHzXVbfOwZD8s5 jtZtU2E7ASlRCuoH/yvRNBSL0tELaDLU+mB0b922P6q5jhtl/Rxp3RdG40dfX/Sz++2Gre3h8pO2 d//aHpf2JkTO29tywEE4sTW0qTEQca1de8PHlsA+b+S4vY1H+LciE6GNzYa6zYDWNgv+n1Sjrdna U5QVMbRtLm3Mfn18MnSEatjxHiSQ73P/1sdt+Wog3w987n23GZf18VCyPl5PI1yYjXBhNsKF2QgX ZiPr4/XjQtbHXeNC1sfJ+jgTF7I+7ogLWR8n6+NkfdwVLmR9nKyPu8SFrI+T9XEXuJD1cbI+7goX OyPr4/+xES7M1ri4sK2P25YimNbHPRHz+rgpAqG/PRDyiERojadlrUQQqfYURF7xYFof90Ts6+O0 JZFaHJjWx08Dh53AoSVw+MfDsj5+OuIfD0HkTo+Wkf944G229XFP5Hh93BYHkq7+u2a8zbReaV2y dMN65XFOf9c8ft5x89of/rRZXdYr6/J3zYWF3Ncr7XnatrmuV2K9bfWA26etHvB2Y6gHrAEuE/40 m/Xvy7O6O/77clx+d/x9ua0ecHps9WDP07bNtR5EiH3d2AM5XzfG65oy8NlzPgiVg8+O8q25boyP xfta0nH29+X9kRK9gNKgdmeg+fDdtmZs+w0C/fcJ9H6G6XcTV8D5n4WdvaAjPGTHBx/b5cv+u4lR aKKZz3RglAHbsSgVGGagx+x4hVk/o+w+bduO1rTr00+rofG8DzBCp7dZYEl7jWCjoFy4UcCln65P 3t/6WPJu4X8v7+98Ngok/g2f9w9BlrwfCb2Xtz5oo+DhUG55M8XBadj6H7zdGJ5vkWV9bkTW49x/ t2Lrf7C54/kWXPofe562ba79D117fKBhta/jGGzt+9u0OeGw78+yPuPFXc8WyeLwjJf6jsH4C/1Z O/S+if7sF6YxgY+cjwn4oeovihH6BT4HBCKUwKvZB+NjcbDDCzn/LdEg8xNHJkLN4z5YCZ+o1nNH YlDNfjfGjo89t1UAJeQ/HHaMAW677bjhYx/BDh8n3PD40BfGBjw6TEUpZnZTgd8MVPO3dGFW2H73 FIXujRdcnpFie9wGUx9Jodp9JK63r+CAr8Ayh08TWOpWR00X6KgAgSAygLGfpD9nxtE8nvboj1o8 mObxG+BAKmR+Ej43CizxTlKpwGOjYAN84m22eTw+hz6Pr6sueOwwwo5XoECV8JlpHTuu+a7kX/Od e980+c3XUif+fvc08fdLE1zz3Sj4zTdNgLfvhyZMZbH1SbYumP58Icw/Dll+j5oGiW0KtDzHDP+m 8HHxl6IPAjEyqMfFGF+KuOhJ73vo5aH/xpn+W+aCRjAu13ruFGIfl7MYnv9Vl7HhRORxzmNDfZ87 Zf9bZvrvehtDPdSaH1nH6AtTHF+fTXHTGG2rBy7PYavv/Ij+u16eg/11/X05HgffBX/Wg2MOA39+ xO65YPhYhJB9HBwFV2tp5t/8ToRREI+DM9Bk2N8Z0A5ZftsaS9tuYkUQ7Ziz6zl8ffkvcOkMXK4G wLWPHU98bLGQ/foSX8/NA37RZoYWfjZe9pzY+JRAnm9Ans1Bt2t2fPCx0xz41PxdPkKP0vh0t9PN GR9cBzzQ5RbkWQ6fT4lq8sHHBoos/a8jPjidkTDbGg4xcN59AQMYtGCao9j8hGncEaDaYzF+HuZR 6PBbA6ExkMAn1udhJkVNEY6MGiAcHtVCmBDFE3aLOkm1jvqGCor6kLoVuYI6FTmL2hf5LLUxchi1 NLIPlRrZmeoX2Z6SRLajbkc8Qv0eIaV2RvSg1kUkUHMjRlFjIiZTfSIWU60j3qACIj6n7jY9Qhmb VlBnmgYKjzftIvyt6VOABbD9jvBMU53Q2PSk8G5TJAqIaC5qHdFL1CdipGhMxFTR3IiFonUROaKd EetFv0e8J7odsVEkidwk6hf5oSg18h3R0sg1oo2RL4v2RaaLTkU+J7oVOUgUFNVe1DpKLOoWVQFl OQxl+gzK9qoQl5FpbKTPv4X10BbPczaB0z8JibwPGj9j/RunT8KThBugvFzydjY3YOLANM9ZC3l3 g8R+g8951r9xWte0m1nztU3nCfE22zzHfl5QH01+bmq5Z9IZEnjDOvf7tekagTRqDePcz531URls mY+bgi1zP5z3reAAwa3g6Zzydkd9/BFsmYtHhdybd0aFpAKHjYI/glMFePt+1gdeZ6Bfo1jWGfD1 yf3TBK8zYE3wOoNNk5aRqcBho+B0RKoAb98PTZxdV9jPxfF+21x8CtTlq0DsAAz+v3lb5gTTQ456 7wvDUHtOD8E46s1FT0fXFvR79I7WTRrDfHD9UMs8C3/arC7P5W3o+aA9T9u2K+sm9LWSxqC9o2fx hj/teC6Oy+zOuThOr6GviehrJfT1K0+7OPT7yvVdW8F11zUUoYuQyb/wOcCr5twOH+viZfF9R3O7 p83HlOa7VDg9+/l4EGKf716HvmY45NMbCvGxHQd8bL0X1/nuBMb5biziPt/FZQiAcm+FPJsDHz/v mnzwsdscNEmAGS/+N8zKpTsNbNcA+NrpAJQ7FPKeDBxm2nHAx8Z7s187pcP1yHwoEb56mgL/4yuo dHTvmsSmBZNO9OPO5ud8a75MY5F92Wzj80rg/zOcsC7k3u8AVsN48nrIr4z3duh5Oxqf+aimtWI4 lz4+z4S83wEOn8OnwXr9MjPEIHo95B3R5/CJt9nGZ5w20/jsiibOftNgPz7j/bbxGT+rGo/PeN5n G58/CDzqfSsYQ+35QSAG8/hsr6ej3zXQ104aw1hQl3Ur21iA7X6NBfW9L4O/0NcMG4P2teZA1nF4 sMzxOOyudStX7k3Wdw5k11/UuF/sKA7bOMy2noXHwNawcyckNgf8OdhuDMTH+JzHQPfc8xkDOyMh z/7AZ64dH3xskpd77vnY5/0u4Hfo21SQ/reQ9592eeNj38M+fyd5TzGXYSKaZL4XN7V69Sza/OuP 1BpjHP2enP19wyao5pjINo8ZCTvPALehwFtmN2bjY7He7HVo4Z1sfcIBMxc6V2dzCBNoNQjy3AB8 PrDjg4+93kjmEHV51s4CKNOPcMKL8LnSOodYHPyK5+LgI4xziIZ41s6zkPd64KCBzz+sc4hng/8Q LQ5eL9LAJ95mm0M4e9aOozWSxvAbkrqsVblrLLb9juF+XJfRtb9fzzyq61hcFdLw61OuvCeovmOx /e8W6PeI6vIMGvy7kbMw36eg01NCAuk+NftEfEzhw/67kXvXu/hXe+nm+rZ/T43tGTS2ffbPOGLi Vgr5L4T8Z0Pkcjtu+NivLnPLMI/lzrjZPrk8G4ft98xWa9S/Z3bXfTlX/MAdv2eu772eXYA5AQjF Qxv6BcYrrV37wsfe9rGc56h9pcH2OJiZpEP9jjP/ljYF5pnJaLJ5djAOMBm9ALVOb/e25y/R22An VPs3t1G0/fi3r86eCVaf3+nMibKsX+1uBnNc6/rVgqgk4a5mSYzrVw3xOx0t5N0FErsCn3Os61fa Zl2EV4CDttkcId5mmzfY30u3+WbDv0PtOLp+yeKbn5bX9M3d5U58E44LLB8m60zLM36cZVwYIdyL 2sB23LjavmnLg7Nvllt8sw184vSul9fNN1v+VdM39/xV93eoUaimbzrqX135mwWrkb7WzX0tl79Z ECH2ecY7zS3PuuscXftZd/gYftYd21g+0Hz9ahvL8biOzH8fQP+NJ30sZ/r7Bbax/H69+5KtfWFf xXUVN+7+r7G50r7sedq2XXn3JdOcFkeqy/sXl+PjEZa2FhxVu63hY1yeq9jffI2P74/MhpE9FjDf /Dcy9LZGB32Md9Te8D2aEGteAbTtxmICJEcT0PNoCJR6EviXqxaE+DxcZtwfmOx8yHGe9+Yro8Gz 0s33pix37mbD7An/Ij2DY/5RkL+JZlzOUUHjeOlb+/xtCswCBvRfmzu3h6DG8TUXnntyzd/82xDr thBmerjMOM9Z0N5cZdHGWn5PF/I/jdczBlq2l8D2i1gPQDbgZcBSwDLAcsArgBWAVwE5gJWA1wBq nqWPXA14HbAGkAtYC1gHeAOwHvAm4C3A24A8wDuADQAN4F3Ae4D3AR8APgTkAzYCNvEsfyPwMWAz 4BPAp4ACwGcALWALYCtgG+BzwHZAIWAHYCdgF+ALwJeAr3iWv0MoAnwN0AH2AL4B7AV8C9gH2A8o BnwHOAA4CPge8APgEOBHQAlADzgMOAL4CfAz4CjgF8AxwK+A3wC/A/4A/Ak4DjgBKAUYACcBd6HO 7gDOwPZZwDnAeUAZ4ALgIuASoBzwF+BvwGWAEVABuAK4CrgGuA64AbgJqARUAf4B/Au4BbgNuIPz BJhwBwmOO3pq+lRldJtRw9tGJ8xKnjt1xmS82otNZPERaBS4z4oabJB4Q/c/n2eZlwGT1EtXBOEX rwL5l5Dp5lXTtVsVHryqK+iVrEtXwiiYNUTD7DzLhPbs2YPOnz9vQj/++CP6/PPPjeitt96KQVlo kmnSJDR8+HDUWdW5M4qIiEDCluUV4cdNlHf1RlJ4eYXpasWu8ooeF67F/3vdQ9xUfOrfm1cuXWl6 puqObzloMAqu703RCPWdNTV5+qV4fmDZ9YAnhHw+0qGnoDypC1A67I0sux76gtdw5QyTMmN+2sTo kcpPZmUsETxSdt3vCQG6m8EfOf+FCcrpqOuo8grZhWuHZVdNlOiPS1eiL1wbEx8B3VVTXgfUUQyj SoSZkai8IurCtXcxo/IKzGdo2XWqV3lFgplSR96Fa1lXrwwQPyW+euXkrQ68S1emXCq77l9eMaK8 go9GvQzF6mG4ncPL4Z2/wftsNdSEoB0ljKA8NpdXeJZXeJVXeJdX+MAXKKJfecWaM1XSDXv4STwe 70uegcdXoAk8dNcriCcOu8MXhwkmQK2+fH0XdenKK7cFcQtCPlMMOrUgW7ZDEH/+hreOx+8SHLCa nxAWFBobcLZZk8lBgpFhQcHUBP7mwBDhojDRojCPRWGeQe8FLt0NSSTdprZ5XwiP92q+OOLhJm+H BIRuDlq2h+JfuvL7bcGw4PKKwdKxrbsGl4xteVIWPTf6aoWx74iz4ZODx7ZeEcxb256/tv1qwdqH Y6SvhIhmwS6PjzJb7w1dHOq1KMy7Q+C7/PmeJ28LfMRfCLxCFocubIoWh0a1XRy6WbI70lsW1ckr KMlnUZjvojC/RWH+i8KSxIvCAhaFBS4KC1oUlh/cIfDDA8Dw71jZnapp/K5Bf8Lgk4zubu7Tj5cw gJ8wQJAwgGqXMECYMECUMMAjIT5PNvizPj/LzvUaHxvQp19gwoCkoIQBwQkDQhIGhMKXsIQBTRIG hCfEZ0d4DgmMkXcN+uAwJD/wtiCrc9/Hxzza2yv4QN/gYOVovrdytOBA301HqZTQwaMmj5m68zeI V3lbEB09MiznoGJkmKfnn4oqxc0m0fGBYc3Ga1usbvJqs8lBVNz0ZsLVUaIJnqu6zRw/OWjU5adW BHvFTfeOm+6zusnFouOQytaYMRWy5cuClwd3k7Ud59VE0yJmxaVH+P4LBf4ZgybFLhQOjV0oil3o ETt36aCvT1Lel648eZviNxMalb6R7wc8NGI4L3nATxlpfXVnITHeTzEKSt0ltI98hHyCPEMev1T+ mWyP7CP5WdnN7nsuUHsvXTn+Fbo5+yveyJA9su/LVh6ueu037w2Xl6cn8eNWn4tdJ1KP0ISoW6t7 qNFQ9Tj1DHWW+vUlAaEb1FuyEycHBTRbtyI40PfToH+WDl0RHNz+0xAf2BE67NOuYb6fNvH9NDxa unZF8MdNczZ5BUUsCovsEDg1axf1Fe/fK7nUQuHnXieboB9Q7pmq3ddUEwMeRn5xca8n7R8W9M/U kbKJMmpul8ELKmaHrOx0rUXojjlrkjwz1nplrPXOWOuTsbatb8Yq0Qdxr4c9v5vaqw05qj1TtfZM 1YG7m8ZKmxS/HdO53d5ruh1gvIBOCculb0k/ld6NfrJ9Wrt9AlnwkNQdO8em7Ni5vEtAaPbX4Qf6 TttDHdD9oTtTtRydqYoeMzmpQ2Ar3lfCjuKhhqwRP33Le+7EyVtHQrOFHXht4ld7TA5M87t48Csv o1fsbXGAvyQ65JV1HZt26CO5O2ZceHbvnT82bfFHs7NRvXd2bfs26ifNM57+Z5Y0WzpWP13/pTRr puylTnu2fq29oL+gO3dzJKL0S0xt9L2+iw0YdHB591Y8tK+wp+Gy6VmD0tApPFv4LY/XcrXHpdl+ v3zLC/P6Sqg3oFOGawahsYmxrbG3cbhRYUQzje2OHtt/wtRfMiRIFvpMUySULP/9T96hk2//sjbp Z59nfbKU4iXiNeJl67vGtBF2bZGk7BT13czQ6E9lIdKs1q2nS/XGN39q3dp4Nrk9n5/cXpD8cIcj uUe6GuIld0xILF5kCPgp/MkdW/OOao6IvkDhT74x/9vfc4+E990j+UlyVtIyXxzw0yD/lp/8jjLz f/ce90zLJrd4E6iZ3ZpcFnsiSZAku3ymvzS6xaums51H9ZH+8jf/rUOCzYeor2fyPtz+2+rApLgt Tb+81oY/N5fHOzerb8fAI/KYpLhHk8rkARM0R2Yf7JuVy/tCO8D0lIlv/OSxux/v/CVo64WX8zOP dKLSt/ttPFgy+KVvPolW/LPzJ5MnTHvGjRuVZBo5YmRcXMdxyaYFHeM6ZMxOTzPO6hzXHbpWCplQ HpruIfiVZ6J4cFlwyVuZ6ZkqgQnbbVNzHgyLwTCJj+0abZoX3bHjQ9HDJoozMibOOm4aIw5A03y/ fg56/K5o3hoUrUMj4Aph2Go08cJC8Xbxmaphb8LVR0cB6rVH/Jf4rjhA8pCkiyRroORpyfOShZLX JO9KtknQosHGiuOQxb4fAv6awSsO7OJztvKmF1qGzvrOE9+RTPG4dCX1/fIKISXtIjXchuE0gPfw uJixMf4dZkuXSd+UfiJFRVK99JT0mlQoayJrK+stQ8NlChm0Ktla2SbZF7IfZOiEzCjjyYPlreTd 5UPkz8jRC3KVfLX8A3mh/Dv57/JyObojFytiFJ0VAxRPKaYqMhVopUKj2Kr4VvGLokzxj8JHiZop pcq+ypHKicq5yqC2M6RtMsRfKMMf7lResW6P8txN1IUX0YqnCla1UnVXfTb1M2XiFA/BSNW5m4mz 0oYFIz+lseKS90JVgUqnOqJCI1Q3Zt4ZBiMZCkTt8FwDJhbpaUoTjK+4ukaarNW1wGStrlkmqC7U sXPHWFPHTt1jO/bsGSvp3rnXsBHH7wq7lFe0gamAOMxTGF1eEQKboqRQ8xi8BIbTC9fm/Xv9ytXj MPOt0InFI7qIQWAYjgeWXeeXVzSzDMcTec8Krl6Jny/2EazwO3krGQ/HYz07RD/HHwUDsed7Yq1o M/Sd5RWvn6l6qbxCYEJJfb1//SDtMdPY9vF9Qz4bdMqwIDcsHuYFe2Rh4aG6IZGUrF2YkSeTVD0p SZGc/odnEiBqE0xz/4I+eDUKjvc9fyOvWYXss2YfhPOE5fzv9IIAFLiFb+TxqA5hz/Go2clo/D7e xaufCO+eChKUhQmvBAmHzPaNvtV04gUoXNR+j3eXBsyOXBF+SwgjOAruBtOBQRevDv6LSohZ0mIa v3VCjADNjVkoNr0UUHEN7Wh7ziehlZG3S9hVOkh6puroK4Hnb/yzLsKwI2S8tzB+T4hxR+igbl27 don7KSJzj+CK5KRMeLeJ9K2XO0Y8x+NnVAXOjYlWdQ8euP3H7K15XdMFGc1jNnfaMd7nYFxScV5X lP/Ye3ldebruqibTvwmb/s3ppXGHTgV3bMWjPFr4DRROb/cXBdOhuOf4ST/wfuBlnf5H2lfat8wr mecbcytGeLq84pUzVb4xoABPPGDFivDx6yQXhemdUGWzvemdeM06nAq6zJ/R1xtFnArinRrIe7XA 8+2+6wehQd0+kXcI/lbOa3Grb3ErDwXq1yVa0Umx5KlbfW+1GvvMB2HRMxVF8rXPzuh76UqnzxQX hfni5/izjLy06LuRE3jPxaTADMm/wLPpfsHd8zf6K19IS54x3xSdoEyZ/cLEGcvg0iol4S/pXRAr 9+qZqhaqhI3eMCLFJSV9M2F/eBBvYfITisH8le3570hvpCbzBIvO+RS38kwGhRcmt9nvcSrw9D8L k1ctnDp1YjIvsmdXI6KSh5qeNryffQPFxxum8Us+b/1Yx183GF/M+LttYcDIm99tipuoKgzYkH4y f/bJvSndsl6QqWSrZR/0HN3cMOXi8YRtO2TRT+gEgY+/Pee7w7JoVZuLVJ+qdt9ekumi83o9kbeM 3y921emgZF55cDLPaxJv6ou+1CT+4qX81BfPBQW2yFh6Lmjsiy/yFi8VrF8RvpunjNKEbFjMR/yD MzfM3jQ724gOz+JnKr7qePbZASmzRsx76Vs+ajlgAq//GsHQN8efCs5Yeqn/uaB5L/LyCjyXXPwx e8ucE+pP8z5e/1NEfFaA5iHNL8/8OPb0nO+fOJwhQZHayOh9aMjHyjdefDEsYfHSM8Jl8qczoXph kPGbqXn3U9kG7ZaNF66Fh0w7KduHTuL54UWq36lAvxfbzEydmXzx6uIFyoqFEwe+e7ZyHxKgUYdV +sCT722NOFV6p1kffsBdk9eTTzwadiqA//zLd9/2vKFcpXtft11XrEO/6S7pbuv89dH6TvoE/Wg9 mqJfoM/Rb9Bv0e/VH9Wf17ep0nsbogx3rlY8lrY9fWvCDGiHHoeoeUvSYnxU2SPOvdczOTL5VGBg eUVQeUUrmMmGlF8/PLNJ74QBye00HSNnpDYrr2jedPs+XlJMecVD+alwFdKqvCKpdXnFw30y2pZX tCuvWPtIeUX78ooO5RWpS6Tt4hI63hnVubyiy5KZ5RWb58Z0h2l9eUXP8ope4K+9yyse7fJba2MP 41AjGmecYcwyvm780LjDeMD4hzHtL+NdY4Cp2T+JSBlXcSLjWzQAzehSMQutnp8mGWk6dzPshXjT VO+rT47aM9LnRPrlyJIn0sP3J69a0YSH+sdVmPCDU/Ctm8DnnzXpTX2HBlXELUxe6nWdLw4RD78g kqPp6JeJlw/HHRmf3KViKoqZ0wUNFD8tTt0aURHXavbTt0w54p6zE2a2FQ/pUvHKxO0VvXkvJIh/ F7/7E5/Ha/bPCRjOlrxXlaQslBjRYMlYyXTJYskqyfuS7ZJiCfpNcklyWwJjvLSTNEE6WoqmSBdI c6QbpFuke6VHpeelcVVSmMvLuo03CVqPf+TKbdWdQUdG9Js1O31K1MCf5LLrPbd9RM2WLZOhN2XP p764ZIfsgOwP2bQJTTKRSRYobyHvKh8kHyOfJl8kR2r5e/LP5fvlv8ovym/J/RS9cn3jP0L9hj/0 ESU49X3fmckLD6e0589LRkjyVxqMsz3Hj3lyrg7NUmQr1ik+UnypOKQoVVQoEF8Zomyt7KEcqhyn nKG87oUWKDcotyj3Ko8qzyurlN4qFKWKVcWrElWpqjmq5aq3VOgV5deqw6rTqusqkTpc3U6NHlU/ oR6vnqXOVq9Tf6T+Uo0+VpWqK9R8TYimtaaHZqgGjdPM0GRpXtd8qNmhOaD5Q4MOqu9qArQPabto B2qf1j6vRQu1r2nf1W7T7tMe017Q/qtFFzXNdXG6frpRukm6eboVOpSn+0y3R/eT7qzups5TH6FH 7fWRuhH6CfoM/VL9ev1mPdqtL9Gf1F/VU4YwQxtDLwMaZuitTzO8aMg1bDTsMnxvQMcNlw3IGGRs aexmHGwca0TTjc8YVhnfN243Fht/M14yottGf1O0qZMpwTTaNMW0wIRyTJnGLaa9pqOm86Yqk7cY RYljxfHiRHGqeI54ufgtMfpU/Lbpa/Fp8XWxSBIuaSdBj0qekIyXzJJkS9ZJPpJ8KUGHJKWSH8V8 aYi0tbSHdKgUjZPOkGZJX5d+KN0hPSD9Q4pwj/u35CFZF9lA2dOy52Vooew12buybbJ9smOyC7J/ ZchX3lzuJ+snHyWfJJ8nXyFHefLP5HvkP8nPym/KPRURCtRe0UfRQT5BkaFYqliv2KxAuxUlipOK qwpKGaZso+ylRMOUzynTlApFrnKjcpfyeyU6rrysRKogVUtVN9Vg1VgVmq5arFqlUim3q4pVv6ku qdBtlb86Wt1JnaAerZ6iXqBGOWq45lFrVEfV59VVam8NitLEauI1iZpUzRzNcs1bGvSp5mvNYY1O fV0j0oZr22nRo9ontOO1s7TZ2nXaj7Rfavcc0pZqK7QGDV/XrFCCyhDixW3psi0aIQ/o9prpRuom 6vwKVTr02lcv6LbqvtX9oivT/aPz0aNmeqneRzdSP1E/V/+Kfp4OFejzdEf0Z/Q39B6GpoZHDOgx g9zQR59ueNnwhuFjw4aDvu/vOWDIzj5+5Iefd+jf1iNfo04fZ+xnHGWcZJxnnGxAecbPjHuMPxm/ MfxsOGf43Yjam4JMI0wTTBmmpab1pmVGtNtUYjpp+s3YZRvvmreX0YgeET/671BxJ3Ef04viXPFG Mdol3mQ6Lr4sRpLvxTxxsBgNlKSKh4gXSl4Q07qG38W4awiFPmOUFN0pS5N2j1oo/UC6re1bkla3 UJz0htRDdlPyiOwxmVzWR4pmyiZI18qWSvFUdre0RIouykLlD8vDZI/Ln5Ur5c/J0Gr5NpltKntH vqoZWtpsnry/IkmRIJ+veFWxQI5y5BvkRxRnFDcUHsqmSm8FilLEKlKUs5WpijeVnyiLlKjVzZZK rexfpa+quSpO1U+FRqkmqZA8R7VBtVK5V3VUhb5V/qIsU0ao26v7qEeoJ6hRhnqpeq7qU/XX6gLV afV1NRJpChQ6xavq19RVqjRN7AyUqcnX7NS8rz6hOaTmaYO1qJU2RDNI20MzTTtOo9bO1aKl2v3a X7UHNHe0f2lidJ11aIAuVvuY8mntq7rT2hd157TIqPpa163qku6ato2+iW6YHj2nT9MrdDN153U7 9Jt0f+hRgP6I3qijDMH6wYbu+ukGtNiwyqDStzT0jvzScNbQWYnuyN6HK8Roo9SQapxjnGhAbxk/ NX5tLDDoDEcMlOmGAXU0PmpqY3rMKDc+ZlpnetmIvjTJTOv2nTMNNL2uv2p8RIzCTEZTini2eJk4 yfSJuAjc3XRKfNh02rRSLBKHix8Hdxfni1+SrJVskqwTz5acAHcX86S3PW+J46T9pL8a/7qEEiQv SReZPpZ+Jf1ROjT2dyniy65IWstCpQ9Lx8kelz4rRUrp+7I1Utzd/yWbKtsjQw/JA2UD5VJZgnKh fL9khgxNkq+XHZW/Lv9X7qtoroiDaZ+io3ySYp5isjxP8ar8HTnSyr+RX1eck0co2yv7KEco0QRl hjJQuV65RPFH0h+KNQoUMGO+zF8dqRIp+6pGqiaq0FzVK6p5Snx18plyj/InJTqrvKn0VPVWt1cl q9PVL6vRG+qP1V+pN6sM6mD1z+O65KCrqgLFDfVJxWTNRMlKjUaDtmq+1fyi2af+R+Oj/Vftq0HN NXGaUO0ozTLtPM0KjU6bdUR7RntDe1bTVPeILnir5N+Cd/kf5UfjBQ/eqK8kQ7VH/x6o1Y8O1mbt Rm/AIHBB96/OV99cH6f31aFR+kn6JN0q3Xzdp/rTOq0OfaP7WbdPH2Fob+hjGGGQ6VGGYalhtn6J fpX+2OUd+gMKxDNe1SeqexofNz5rVBqXGFGaId9oGwLMI4DAPARsM04zLTKpTe+Z0CrjftN240XT Q6ZLRjwCoH+tQ8Cr4nfEWvE3YrTFdE78h2mg1wl9km83yUARCmsaIJaIJ4vTxPmSnZKDEvSn5G/J cXGgFF/qz5b0ktDmB+HSX6UXpeiW1E92W9JR1lqKb+sIxOgl6SclQ2TvSA/LvpHelHnKUYS8vTxS NkIuk82SZ8s1MrRZ/qbsE1mRrEJ+ShamaKNAvRTDFL3laQqF/IS8t3SDAq2XT5S2f/W0wld5R95J maBEo5VTlAuUUxUblCthmN+qQLar2fYqfDGbqkKWcRwP43r9SdVVFWqv3iu/bopSn1A9p5arstXI Mo7jYfxHVYo6SIME6lD1w+qxmjvqxRql+kMNsozj5eq/NH+rZ76xXIp4x5ZpcqU9NZna6RqNljIg 7OAXtbe0frrbmo66aC3qpF2rHa2don1bl6PV6bZo0Tldpc5LH6nvoI/SPamvUqN0/cv6N/Tzda/p PtVt1B/WIaGhiaGtobdhuOFR/UzDeD1aY8g37DQcNHypP6RfZajQI76hu3GI8RnjC8ZxhtXGD4zo rmG/8VfjReMtY4iKMqk1CN8xmWrKNK00aUz9TN+a0DbjOVOlyUscKfY1HTfFwQ7TXPEr4rfFBWKd +DPTGTE6ZNombyJpK+ktGSkOlCTFoqtiy0huG8hRiFQgmSQZLB0rnS59VqKUoCUSy0huG8gRHslb S9tLF0IXPlc2LA99I/tZdk5WKfOSR8o7yKNk6El5ijxRtkz+rmy57C2ZXo6+ll2TCxVNFG0VvRXD FY/K0UzFS4q1is8nTXxS+9RmxY5xaL/8jFygDFU+rOypfFz5rBIplc8p1ijzlTuVGxW7FD/DiK64 rGih6qoapBqjmqZapEJq1WLl56r9ql9VxcqCWUPeRLfmDXgnwthT/bj6WbVSvUSN1qjz1TvVG1V/ qv9Wm9SXVQipg9Qt1QLNNM0ijVrzngZ9rtmv+VVTrL6l8dNKtP4aFK3ppEnQjNa8qn1Hq9V+o0U/ a89pK7XnNZG6DjqZLmxLkmTBu4LZ+dFPbR3VUYv6a5O0k7Xztdh3caUjS6176WT6FP1sfQctuqx8 Q1+p/FR3Sn9NjysdWWpdYXhC/5JhrSFbjz7R7zTXOs8YbGxl7G5EllpXGWcYPjAWGr8zokXycuMm +d4nvjdafBfh+VuOaYNppdEyfUO/GL3FUWIfUzOTZfqG8PztU/HX4gKTZfqGDovbiLOkXZqHS2ZK cJ+ONkm+kPwAnbpRUioOlraSou7S1pIekqGScZLVMHgXStF30t+l5dI7UrHsrqSzbIAMPSULl/kt loiXSLfI9sqOytB5WZXMWx4lj5U3kyXKU+Vojny5fK7sFdnbssPy0/LrciRShCvaKR5VPKF4TI6n 6Ig+R8dTdHRvjp6lVCo+VKIdygPKP5R/Ke8qA1QPqbqo0EDV06rnVQtVr6neVamV+1TomOqC6l+V r7q5Ok7dTz1KjSap56lXqPPUn6n3qLWqs2p0U+2pidC01/TRjNBM0GRo0FLNes1mzW5NieakRq+m tChM20bbSztM+5w2TfuiNleLNmp3ab/XHtde1iJdkI6nRd10g3VjddN1i3WW61FUrKNfj8bo0Gh9 zetRdF5vuR6NNcQbEg199WiOYbnhLcOnhq8Nhw2nDdcNSGQMN7YzPmp8wjjeOMuYbEDrjB8ZvzQe MpYaK4x8U4gJtTb1MA01jTPNMGWZXjdBJ7/DdMD0h+kv011TgPghcRex+bLuefFC8Wvid8XbxPvE 6HPTBfG/Yl9Jc0mcpJ9klARNksyTrJDkST6TWO7gop/FntII8Nk+0hHSCdIMKVoqXS/dLMWzr5PS q1JKhq5J2sh6yYbJnpOlyV6U5crQRtku2fey47LLMiQPkreUo2DZYPlY+XT5Yvkq+fvy7XJULP9N fkl+W+6vwLdpEhRotGKAfIEiR7FBsUWxV3FUgc4rqhTeyihlrDJemahMVaI5yomJK5Ua5ValZfK1 5x+lj6qZSqrCw7Z5oR5fx04creqRpXpdtS7NG/Gnjl2n8j6oEnyfFvI9dOmUOkyNIrI+VT2mlqst Q/kbatQOhvIf1Qb1FXDuUM1HatRT87jmWY1Ss0SzxjxlQwc1f2r+1pg0gdoW2q7aQVo0RjtNu0ir 1r6n/dw8ZcuydOkSXUddf12Sbvrm0Vv4o7eIRm+Jm7otiRc+ekvm1tFfpG1N2zp+7vOfbz4+kxq9 RTh6yytHFgpkKAnxxCvyto/e+l78ns9/2Xp4R0pG2NRts3dM3zI+enSh8G7FNt7B2cfT/v3i1anb j3zx6wYZvklUIYMeY9dLFTJKH/bd8Zkt9d32jFTd/C5uoqpCBk1npP7976K++WTrEd2hPTn6Kh3a oo/6Rrv1e/0Z3WX9WL388PzBH+6acoSHRiV8tB0uL7q1zXv1l+Mz3zyx9qtRxfs1W6es3xU1dduo zK0jive/8cWPhaO3vj161z8P/bHTIIt/9Sfd7e77jVNaX446GXZuxLPz9wge/Wkt1e/3kx/vFyoO Tj8VcX7qVn5G/yMvH4nuHmww3yV86ZwgY9z4b1af3jHeZ8vJ4pfO/YJy/n7vpXO8jWVR3yXNPHvy 6Md7An6fyT8+s93Vn6jLpwXXzwunbNVsjU/TvajL1W3U7dJ9v2v8Q4Y9XQwDDYofL5/+c1bSFW/k ffk078BNXpR+yY3Y86+b0IfQcV4wtb+E13136R8TR79y82nxhyYvw+jPO60V/4+t9wyLWvvev9fM ACKKAUTEmsHeA3ZsGTt61ACCXQPYFQ2C2DWA2EtAseMJFjz2AIpdAyrYCVgQGwEsiEcNoNghz2zG Ofj9/Z/Ll4h6uWfvte57fdZNQGgZNjMpeKDxG6FiidZBY10bmX4vv32vl/NgVHKl6afOXawf4D/f PwD3D5kKM5/MR/Jn/ahzMaPONZ32edTZwAQ7Bvl+hx3SyCnBdkGPvEoCE+c/10bK2+XDsn7Al2sv AhMvaCY4Wo+6FJh4U5/Wt0xb5DKw+F8rA3QxzGpmE9wMSsrLVrYIz9/U0tyKbc0bID4b5pzV9dRe Rf84J6K9pevNs4q+j7hwyDRp4tUsRd/22NEebgUMNFMKvEpOf3s5pIuNR1cbi/4ddcFDulqO7BrY 1cZy9e1z0M5+ZvGXJ33LkhxqlTVrqS/17EQMJMaqaYnc7XMJb21gW6tVRDPKWNFbNbNJdfmYP16r TdYGaK5a6u60B4FYOuh20/X9ZmT5lqTiD/Cw1/hXvDpxO/v5vP5uExxrZt5jg9P1zUe2rzWlmdUZ tfn3MM1n0nBJnvPPph3LluhDa/NTnBpY5ZV/3eKTWVH+1XP2qfeTern3XCpOnbdW1z6s/0DPn36h tnZ+Fq8CtAJUC7W951Xj/ijd+DD7LVDRZGai1bMxTnfGhDl89KqboMvyuuNdR5ftZVV7sibP+5l3 synO8VQqsXCITutb8q1o/rQNlntvn2s8pfUc+KgLq1iirxn4V6lG900lDYuejWhALFZsDBq48OGN rwcTtrhOUgtSv7Ka2/cc0m31h3cdpy2J/dn0YimhH98zOWmlzwOyq09p6YP500KmW1n372TsE1Zo 5k9bttBj+crpr1aznkGrXn7jHF2XOMon0aRccADNko9WOs9iY6/uObA4rFZZ19H60oH06PEDy0pJ 3+aaGxfbvMU0FtqRlPM9u1Grtk5iuEDdiNqfc0oeU1BEFUYEJpZtGtI2JC0w8c4jrVdJq7F1P+aP Ou9VMrne59efAxO1RV2mtNZrDAY2xyJ2+nBDeaNXca9SctO5mF2e8rU8vkVY8RFiBHOPWF8Us3Po sE/ND72YFv92hM7+r697FtyoSeqLZw9Xavwady2X0efMJN481t17l99BKXAqGfDtpVNJ9Y3BTiU1 wzV1StvM368dtV834ZDltEPr9o+bv18TdEi38/a5S6O5hpstov7WgvZitGuafMPg3T+y/3TbmKFz Ne1qB8860KplkysW7Rw0XU7pep3/+DJh2qGsT+v2z9qviX35rcggHty/XRL+vrRrTFwbg1bMcxQv 1RhD6bGrlhaheJit9DzqglGHBR0qfxV0yPLRmq7Wg6d96vjq5XEyIUran853OnU9j7wAeVYueUpf jcXH/OeeF7LHOFqMsec7rvDHN1p2T+G7XQAV2cHIDT6bWD+f7wV9bhRX9Ferd6/4kv0xb6pvhUse 8F0KuSqn67kM/+t0BSqwQvnT6fql2NRSy7dE9Q54ONV4DUJS7avd6Txrn60ddPSrEfE4NVNvr7W/ X/LBEzA/qG3DuzryrnV4V8tUR/8e0Y7t7Hvta+Dr2GiHf42HFzT7XHjXUi/Ovr9bc961xVHetaVr TOuoyLa8a7vDvGua16h9BO/qyru6rX9l0Yl37RwSzbvmh3cz/sbuvKs779qDd+2p5V17tfjmjLXB emEjMD8MgrFIbDt2GLuA3cVysWJsghbH6nkD08Rufkv9woEwt4VdMCxu/eJgf5zv7pRhKFlnUz8m 5QffaRDvnvfbFx6Z8NcaDfRvYveHL0zhqTjfw6FxyohorM5hCdcSvNvwYS2oz0FTr+QVu8II8G9h N7M/LGihR14d71a/cUrC99AP+Aqi7bYB8xoQQxxa2E09ZNfeZU5X4g6xpVSn1WD1JEJL5rq5lbzr Aq3L6mB9SE8ygJxPriZ3GkUceYnMIPPIUtKCcqJaUdCDGkZNooKoCCqGOkSdo+JuU8+pjxTQ2HhN r0xdcQM/zGteI/eSSm84ZEb30r70Z/ek3c1gOo2MnTPza246TafT2E6ovqeE/j+CIJwxCYJbzDPm AzNfZaq3M+zW9xveZHezmPzwTtH+1fWTA5N6RAMJlc6wb/Ff82CmbLZ3j7PI3YUqe9fk7sKb1oFc FLefO82lczncOw5+cbV4sw6YwS/hYSO/lEvgr/IP+Tf8N95GgIZCB8EgeAtThAXCWmG3AMeFPXym UCB8FqxEZ7GNCL3EEaKfGCxGitvFw+IFEe6KF4ViUSs5Si2k7tJQCSZIc6UwaYt0UDJqe+mpBO+l Z6Kd3ERG9XGMPEuGZfJmOU5Okq/L2fJb+YcMNZXGiq1stvjWK/9ZfK+UL4q1Wl+Fdmoftb1itviO qr8tvlI0TTUKgB4YDMMmYcN/23vnsNsYmOw9B7wZ3hVHwh6Qsjfbe8jdgz+df2T8A3L+N+EJlc4/ Mv6hIdmBNJDe5BRyQaXkg+PkFTKTFInPpBXlTLWhoBc1gvKjkPbfTh2mLlBwl8qliimZdKRb0N3p oTRMoOfSYfQW+iB9hr5JP6XhPV1B2zEq1ZkZxIxhZjGwjNnMxDFJzHUmm3nL/GDAZO3hDHL2FrHr WYhlT7Ip7H32FfuFtebqc2C29rzY1dxO7igHl7gMLo8r5Sx4J74V34OHYfwkPoiP4Odxh/hz/G0e nvMfeRAchGZCV8FDGCdAoLBCiBL2C9F8upAjvBPgl1BL1IsdxQHiKHGGuESEjeLfYoJ4VUwU3ojf RBsJGkodJIPkLU2RFkhrpd1S2HHpipQpFUhZ4mcJe4RLAKBpktnivh4wNBDD5IHyaFlNg3ly2DNa PiAnyzfkJ/K/MpTLmOKilEsDldHKTGWpArNkXtksX1MeKYXKd6WGCo1UQu2ruipT1YXqOnWP6hxV FJN3QY2MzCi5/OmwApuUCjVRQcLBpBtgrGoWDqfUNPWxCnewxrhFpXAw6QZYjJmEw22sxf1Pul/1 IB9rRHSo34toRrji84nVBCDxsAs3iYdLxCccLInu5CiiJzGXpIkY8hAJJvFwh0DioZpRVcAgSqkW QLXuNZfaRh0cspGEug2aUIXUd+ot2Ygm6L40uFKTaR9qDb2IOkZfpk9SkEI9q7R3reneDMX4M+BJ r2QO0v8w55k7zAtGYSCi96LesxhkCnRlKk2BQGYFE8VcYx+xhex3tobxMjO1WD3ry03nRrEbuL0c CFxdJyduH/2eq+DQpBM684P4MXwpvYKP4lnuNA/pfDJ3g3vC1RQaC65CPwF8hGnCImEm/7eQIPD8 QwHeCN8Enk1klwthwjs+QAT95jniDvGIGCNI4hXhkwiWUl3JSnSX2ogTpRFiuAQzpUXSWemWdEFU pFyxtgzN5W6yXiK4odJy+aE0X4bHUj6fILdyfi6/luor1WXoo3gqAYqXPFnOkQ8ru2S4q2iVa0q+ /EWxVHqorRWYpAapEeo8xUlt3/O4mq1Cc06hY9SvqgPmoo7CZmAwWt2I/Y0lYLyaqF5Tv2BQqDbF OuD1cQLrixH4WhwWYsdxN3ztq8dG8R+pvMKgEWGN5+O+xHRiMTEY30sYm8q/8QfEVfwhzhLfcBsC epMdiB0EmuftItcS00mQyCvEJ/Jjqw9EE6ozdQuDXJuuZCjF4HuoE5RI9RoBd6jP1EvSma5G1aOQ FwAU5U/F0KsokxcwnoYk2pHR0d0ZF7orN5c5S4IfPYZZT6czkb+9AEBmQFMGeQFjmc3scgY4Zh9z innDPmZMXgAgM2Aap+PWcyHs3dC7LKxitZtn0yDY8t/YTvxAHkbzM/ml/CyO5xP5OC6Jg+tcNveW +8G1FxrzI4WpAiwU1gl7hBNCLJ8lWAppy6DF8Vc8zxYK99mx4miSFSFaPCAmizfEM8K/Yrn4XoAK wU5sIlaTBomLpVniMhESpWvSI6lQyhZryI3kJlllOJ+i3Z2GnGHjbT9B9pLSbbtLqQssJQh+vs5Y BJ7K7+UK2U5pokCFNEgZowyWI+TZ8t/KQxn2yafkNPmMUlNtrLqq/VRwU6api9TpSogSodysZbzi F9hP6itlgNAW641RmD8GIViAugMzlQBUASBPLassAQcx5OmF47AVj8DO4oewZ7gjXlkBAH//uwQs JzhiHwGniP34Y+Iu3r21pAxu14qE7i2s3bWEPTGWCCB2kEdIuEjeI2Uyg9BRyEEwfgza/eEg2FC3 KHhGfaBU6iPZlHamEIkEZXgotfdjT5qjrtKnqLc0/KBrMo0ZW7of40ZPYRYwEE3HMhvovbRAFzAP aGsW6rPt2D5seyaA9WIkpj0FUex6ZjTV+NhDtoJVmGYcdOU8uHFcIDeejeJYY5mHA2wye4N9wv7L NuYxzoUDUx03lfFU5T4Pr/jGwmnmDV5LkHhPoS8PpjpuKuMi7yuAhVjGVxPqCcNERQgS/QUw1fG7 4gshV5SFyWdhCfXpzWJxNdVWnCNNEqMl+KKgC/5M+iCp0kexqQwOUjNpjeRhVOub5BVSogz7pcdy kfxTtlVwpZbcX4F3wlRlobJOmS2HyX/LOxW4Kn9VqqsN1PYqqXZQJqvgraxSd6hH1IsqGvBEqFCg fFZaYz2x4RiNjVBXYrANK1bPYrewZ9gHzIo3XvFwsRs+BB+Pz8FZPBrvjEMyfhB7jBfhP3FbogLL wKEJfhCfSSwlNhE8kUjE4fCIuIIfZKqTDcj25EBCR8LgEa8IVMnNhRxQJS8jxpCmDo8iwZ8MIVEl NxdyMFVyZ6oxNdf4hM+koc+lU3Qa/Zguon/StgzOQC26P+PLDKAXM1to5BNBKpNAv2a+MtXZBmx7 lmShAzOZDWXXsPGrR8/btzCWhcPLzjKPmDK2GlePa8v15gCZPp4sMn2OcDvZoyykcRlsHluHb8mb PB9Apk8Qh0yfW/w5jt8CPc9/2Nntck2srdBboATjUYYIq4QdwhFhJ39PkIUSAfL4Ut5CcBLKhIki I4aLsFWMF8+Kt8RzwgdRFe2N9Vp0EJuJyBJcLnHSPglOSWnSY6lIyhFtZVx2k49Wz8QDU3TT0/RD snygqdRFGiyNlWZL6O4+lsF86j8lN8VXma4ALuVx65Qi7m/5gfJa+aqA+dS9VIMSqq5RYYGyVzlS eeqfVEusLgbmU5+H+anbsH8wOI8xzAtsF3N67iUM3d0/BjMs9ns6ewND7Vs5huH/B9zgcdS+wTfi KlGfCKZa9LEhJ5OhJKBHHQ1mJDKfzCQsKahLtaacyTZkL3IEudJYvOEf6jx1h3pBKZSGLiab09CN HkLb0OpmeyKE2k+fpiGdzqHf0b/oWoyeweg/RjdL6U00IjDAhGCYCIw/RjemHh216PBnjx7M+bO/ Rzd3OTOBASYEw0xghHO/RzfveTOBASYEw0xg7OPBhGDUFM0EBpgQDDOBkSqACcFoJ/WRTAQGmBCM KgIDPokIwRgmT5JNBAb8MX+r1KVQWzYJU7MuBbMwNRMY0EkxW/gmAgNMCEYVgQEjVbOFn1lJYIAJ wagiMIyPvHlwU4wjAgNMCEYVgQFniHjcNLixIxGBASYEo4rAgGwyjfhBIgLDlUIEBpgQjCoCA75Q r8n6NCIwPGlEYIAJwagiMMCJsaR7MCYCI4KJYcCEYFQRGIAQjG6MicDYz55mwYRgVBEYgBCM0cEs F80d4FDzFfaE+5cr5zDeha8s2yr8doaDPPhgPpJfG2UD2vFLHNfyF3ndpSjHS/wr/gsP1kLNuL95 QugroFK+UIB1QkNjKRcF06C2mgi7hbZib5ES/cUQcZWxZYMj4kXxniiLJaJOqiO1lMBd+kuaKDFS uLRVije2bGB+0u3lpnIXebDsG+iRqfXItPLIdB1/X+fr7JE5J8vjSUBWQJb3Dp8JDzKiLTwyLT0y 112LZXUmZ3jzw34eWYakBzeyrmb7xjiNvz8ndFKmn97jkWVFwf3amtCMqPdPNox/eO3JnIjpbDFp rBE5xeQXudi6MCPaSWmVN/LLDVfN1JZ5nZSBSkxhrfy9WddkuJK3Qnkn71dq5e/LuqQ8klvlKcOU vsUeHpnVxj8cV6IgY3j3Qy10NTS4tLwsI3pD+Rp506MZHg8nTMpyL9GCz5wsa4/nrSZlObze9GiP R45LdGbdL0IYGAzig/LWr0tmNayV36GaJ7UrRdehdJtF5+95e15behUFaOq/GZ+lnd+lZGGJfsAz udLzXGypmz/BO3+NNiei/SG4udiyHCKqbds5dFjzQ2VXF1tqYquVDeE/jx7CP9yT1+xOtHZSSXZ0 lsbinaVNkZWNxe1XWt07K8vPVrZONpaWqy/l1Et6ID5a96RYM/Z+WPSjW58C76MhV7x8NovTXspJ cPzR1boFloa3U/2sp3i+C9DYLdd2Lg3Q9C3T6AJtPbGGLWbr2jYZV8LYQ5USvvQlI/qVpt+9zzVF fsqb1LLP5bUsLdTojN0djK8NqRoW0aeH6M//UknDwtrIAz2R17rQENqlY2brQk91ejKyQZeKGaOj 1+rauzR9o/cLLbfpWoxsUKj2b6jtXLcaIV11Yfbb7YsrmozPslrWPbB72EAHzq0u8SnYLbCjbqGb pdXpTprwjss6pnUeRARgg5xb67SM/Ucd1k9XZ9C/lnsv5TS2tFjvHPq1i/+KVnmayKkdfpaIbq47 IlNaTlbDv5GG4h66qYVU0q8SuwRb9uAZZannwaQbmL7DGClty6V5eZpltpHh8sPQkKSLWzF9Lh7C ltrMyLOa6/BRO+DRYMfBVq902dTbEZo6g+IajS4a3mxwS7rEsAp727Qi6QFaIBqcv+RBJOm7tf4X XWymDnQaz+jMMXUetwy8bxT5bdtFP9qK+2rOP2vjqPrEaLSfcZsPX30HderR3Y7Z+dJyhEdm9Tsx i4j1xNqASVmG7XOujZ/6elLW5bK9Wa1ufw3Tu3g82Zt17vXtbdGPvn3vUmFZ3c7JYOiiKcTH1vZ0 NMxXXF71cLXJeplnsAobiinkXCyseklMqnbFoD3WZkf0BKYPvvJSWRer5LCU/vIEOujKkZfvq+k7 vNYaZvp/1M30Lw/+FXzCSzMjoE2dCO3XYF35Astqq9tEjKsTobFfrdt5KefSN/q57efQj1rQTpj3 3xQmq/BT7bzmTbd+6NJmd353ey0Ee9ZYebO09+fMCE32St2LdX5Y/2qrcwNarywP1sRib/sWNXjX w73Jxy58XxvDAS4vmev58zs+UO1bhozRk/zyIPdiHWW/uvEA+9WWZWsC3g5OeoDWMVZ3jDtOugn9 D3k88DmY14V0L85rmaD0bfFZ77J2uWvU/b1ZDW9/itzncX9FSsnsu1M9sp5EeTw2/lGgmhzSxPqh +R4v+uzMrFCrR+Zak056vVWXipbn1jJNxS4iDBbHirPF5SIn7hNPiWkiPBaLxJ+irYRLblJ/yVeK my4tljZIeyVBKk9I6Q2601aJ+4yv1iTjRZnpf8nVCyJcj3sv2lfLIxN7PmXcZldIcIXaHpmOHpnh dTwyD0Unu83J6m6/TNOgSfjBRt37Hj/oXuJisNvq5hPo9UrT3GDXwmDXMmtKndZuU9oa7NoZ7Ka5 fZ3nSxjsXA12bga7jga7gZ0Mdp0d5xnsIvy7Gex8uxvs3A12PQx2PQ12ml43bp6X78gvZEVGg3BA k3DzIBzhDycTiryBuZY3/2bxwJ4w90ZeMCy+feKVMsLg7ASW9mdYe5uLU8pSDLUHGeohk7RwhHPa 57D3ozXQv+Ja3m+TtCnMKlYCVEN9BzFK8fdMeB2kHlAN9sM/fKG2eS6fGp5ZaZLeyFszExbcKH6i /mv8Yn0/MWq372YVx+5OHTDvktp8yI28qYOLM0rmPFYDMatO97WahKIg7ADG2t9PhQ11bqedkWWs BNPhdfCWOLjjf+EmzbUVRzwN3MIRT6Pi9kRTArH7hrHEbMKkuRLaal7E5esuuSa4OJQ/TfXsF1wY GjLjydUC4rN7EvS0sCbrk+3I0T4nJvoiuZUw4+jsLeRB0kTTviehgrSjkAAfRI2hZlHLqLebqTjq 6C9DT4TPtuppEeH/0N//aPFk2y3PjP2UyST9kgLv+9qIVXQGTf+mM/6hz9N36Be0QoOGWVPdlnFj UPs+nVnMwAZmLyMwqcwDxtS+A+rfGzJerKl938XCMfYyK7H57CfWkqvLteagJzec68XO41Zy27h/ uPMc3OFecAqn4Wvzzflu/BAexvNz+AlcNH+AT+Zv8E94+Jcv5zHBRegkDBRGCzMFWCpsEpbxicI1 4ZFQKHwXoIbYSCTEvpVg5UJxnQh7xBOiKJ4UXoplRkFeT4K2Um+JkvwlE4FzRIKL0j1JljJEnVxH bim7y/CXjFaxwmUzgQPP5A+yKn+UmipdlMHKWAVmK8sVTtmnnFJMBA78VGxVXK2l9Fd91enqYhU2 qHtVQU1VH6gmAgcaYO0xEuugelUCdrswQISdhOVjnzBLHBE40BMfjtP4PNwPQ4DdeRxMMIaGQDBG N2IIAeOJOQRLRBNheDKBWAwwwRguJIIxRpMzSVhKbiJ5MpGMIx6RheR3EmpQjSiC6kuNpKZSC6l1 FJgMmCwqhSyjqtH1aGhL96Yp2p8OoVfRO+gjNFyk79Eyjcz2UgqZ7YDc9j/NdkBuu8rYs01ZBwbR N4CcNo7dx5romyIWfrK2HM65cf25jux0bjEHyGkTuFTuAfea+8pV56EB354neS9+Mu/NreF38XCM v8xLfD7/ibcU6gqtBegpDBdoYZ6wUgjmtwkJ8TyOTNJrcTf263FE1PkkCMh2g7h4B9HlIvy/7uvf Ve5rDdEgEaLZfYUq99VKdpbbyK69ZLdbXVLGy5GRTNakBzBE6iDxsrd0TX4kF8rfZaih/JAIpa8y Upmq+MjTZFgkByqiEq+8VMqUamo9Fdqq9RVK9VdD1NnKjf0xL2Hv9wjliiq9e6GmqVlKHQxaYu7Y X1gP1bSZNxGDGPWQ+gT7qj5XMRywLjgMxsfis/HleCC2Dz+FH8ZAxQrxqEIdcbsmRnQlBraG9vi5 f6/haLtxHX6FyCSggMjCLcgyvBVZj0BOC1CEP7GMPExeII8QuWQxCVqqhGhGDfyP1YAoqmmjeo1q UMbiTz0ma9LQmLalcMqNmkIvoNfSu2k4Tu+lEGL5hbZmvlLtGECM5bk3Z+l+pInV+E3Kfme2kjjr xuoZ8GWnsyZN93sB7CVr0ug2LMl5cQYWQrk13C7OQHvTjTkXbgMLOj51Uq1KxdaFC+JpDmJ45Lj/ wz3l73AVvMJBE8FGqFep2MbzUQLLw2kBOe6pfCb9L99YDBWgjrhIiGC9xFtvl4urhYsiHBVMnXup YCEuFodI0EOcIx0wnnyEuEM6JD6T4LZo6twdjK18Ro3hMiyU0+kosou8Uz4lX5W+ylBd+Sa1V0jF S0HQzhQZdihr5euypFxUELSTqUBr1VkZrt5XWl9fpD5RmkuwynjUR5QI9YtqjdXH3irQB/PESHU+ NlkNVfXYLhWOqbmYpHbHnPBWeA+8NQbWeBBOYzE4Z70Zu4Y/wmGWwr57jDkRdmovYgThR8CLxoHE NiISP08gXacQYBJ2XcgWxHgSLQ/BL2IkmUweJJ6QV4nHNEbBGFxjvPVtyelUc2oTxVOQSF2jHlHX ye9UDfoHSRgP+7ewQ646YpnBDE0fpNsyjvQcX5hDt6APTKpJ7mNPMrvoP6R5DQYtR/RlwLQjuI7Z w2SwIvM/0rw3G8wdYmHahBvLVrEGei0XQv/gvmFQNT4bzW3k/+ZhE8dzidw17jBfyNUXavDQiDeP zxbyx4Urwrtr+3BDnLZn/G9YfgT+gp/+4QkfMAoO8bUvtTEWgaXiJpEXE0W4JvIC4ubfCk3FmiIp QejvdYrR0glJlLIkeCndFxE3by05Sk2lmR+NV3w8HSOvkl6zd+VcuVgGreKo6OTuCioBqAIAKgHb K0vAwMpFHWiidlabKmPUwcoyNdlYAQCVgE2qqQQ0xlwx6If5YP3VRdgc9cn3IOntL7iFPflypPoB NUn9oeqw7jggmm4uHoYz2EF8KxaPPgb3Ko06FbPHm+LHjB8DYhmxmYgjOPw6cR6vNGS2q05En4zn hCsxmfQhANF0J0iRPEm8JO+bDJmOZG+qHYlGsSup+SQcoS5S9yiZyiB1dCkZREEG7ka3pb7h4tJQ mqejKKiy4X5RboyehuH0AHoUPaPShhOZBBqu0mYbDpXxAAlC2FXGq+1LrVEFNogtYaGAQXXcXMb9 WPjCxnPb2cPsBfYDF8XZ86DlUB1HZXwFx/JhHFjsaEDEpNfnWxJ3uVqCykFHYYdUZclwPFwXEILZ Sijif/IdRFwAb7G/UGXJCAK8EjdwVpKz1EaqKbqIpGgU6JPFKktGEsFC/iS2kLvLQ+UJ8nCJlqCp vFLaJpktGUUCxApEy2OUWcoyZbPyD2O84k34Py0ZeKSMUgcqi9T1aqx6UuUVYNRrysA/LBnoqy7A aHUgdRS7hGVgb1Q4iL1tvApDLK2pkIOpkm/HvmOmDg+KMS3uiA8hzIUczJX8PC7imPEJBxtS3uhD TiMXkevJWPIkCSmkQLwiv5CvifpUJ7IBCe3JAMqLXE3tpI5Sl6gMCvIoibSgETI9aNq3vv1GQ296 yIQx1ALKTFaCCa0soUxkpTsNf9HTGIYOp81kJZjQSnvGRFaOZcAQ/Hzd5kU5m07Id9lctpg1HqUp E2Ao586iSADYwoWzW9l49iy7vXJwBk140+BsFj+W28zH8ZDE7+NOcWncY66IQ4Mz6CeYBmfrhcX8 SSFF8LkvZMfhtnE663j9u30an+t8Nv+W/8HXFNDdhapTjxXui19EsJZS+HCmjbSeIUW0CgV/GnH5 opMMreS6Uh9paOWpI0AE/jTiNHJXBZARZ0etUHpQviMmGts3MBHUqH1zU/WKrwqof5uhoPbtbyVB ea2CiaBG7ZsXZlBDMUD922TsolqbuFHrGGaBg+lRH4ZPMr7pEfg8DA7h5/Db+HnsDvYCU7BmBHQl TAT1CiKK2E9E45BO5BDviGNE3Nwk1ZHoTwLq4BeTG8i9pECmkgkEmAlqtEfXgZxMQSi1htpFHaMu UxKVT2WS8GePjlp0+LNH1zC1GdDS3ZghzHhmDsMy0cwBBpKZG8wT5l+mnMFYF7YTC02Y0exMdim7 iUXzy2ssmAeljTiC68uN5KAfu5Bbx+3hTnAil8W95KCMq8bX49vyvXmK9+dDeAjgdvBH+Iv8PV7m S3idsSwLLQV34S9hosAI4cJWAeKFGP6W8Ez4IKiCvdhUhC7i/8pTSBNPC3/KU/CVquRpqvRAgtfS Q7G63EBuL5MymoFCqLxG3iUfky/LkowSEsBS+SyZIxLmKSsV2Kb8o5xX7igvFEXRqLVVaK46KkPU 8eoclVWjjVIPktUbKpJ15SqGuWCdMECBLYPUpdgmjMcSsWsYPMIKjRe4Bt4IJ/C++EgcpuIL8WnY HvyE8c5m4S9xKMOrEfWItkRvAjVfIQSsInYQq3G0AiMTJWh4UodsSbqTf5ETSYYMJ7eSEE+eJQ8R z8gPpGndEar2HU3rjmDad8whbWmcdqP70+BLm5AYNAlLpR/Q8Jr+SldnvvXTM2gZYhQDqPva+MdO Y2XEJDJJ86cWMWhvvzXCZ1v9GtqancDqJgY5TjS+6LCDPcKeCCL/23MEVMovs1V7jnDY+MZXuewt ePjTZUd7jvC/LvsgAf502dGe49G3QmBRnLYozqooztXq135f56K4WolFR3SJusTDn6aXH2DmWRTF WRbFrauYsuy3SRpJnC7aZxh5akbi5DNfpjg9/bU/VI3z0xfFW1a0WJmkCWWCNp3f8OvglPMZtQLY YtL4WpwrJvN2iEfOMvPOircSRn65oXOdejPhofhG7HJDSOyTCFNEOgWXNoj9JSGxX+JEyW6BGC59 EAtSPLoHlx/4mVrpkfY8jTzSSxsbP2LmtXvRSuyQPKPo4AR139NULfjUSrQuOt5K3XcquUPynqLD Lh0TzqWf1TsZLl7zczz9EI6U3E5Oy86XHnsWz0zx1UlXLR7dyuuVZll6frcuv/6vfdr52anOWXp2 gAkNrf86RDf/U2KrgpyI9oPzym7Wfw1NC29aTdE9TtI8qf9a07swnqJ/HKR+vO+V0ixw3litmtox UWOx4bXN+jdtbSxWFeg2vLHc9ubkm7o2lqsnnqs38pRfcpvzEJ3/Y3/H5FkPbJPQFRkk2oxJ5Cae S0i+/TjnhjxNjbonad6WXNDl2zUueHQ1wFJTkKaz/VAiX/5S8+XdUu3P1BqB81Gv5KrtmNinephB tkobJwfK9dQouSju+6XkW0nZYRYGg9+p76NJQ4WrtvB2ctHt5FU5+dlh/8qVfEqRu2g1L2dT30ud Ff0GO/tx9qkIb+/zZvns+y032b248lmntdec5l7ucNDtdbKMnHiu0emRp19/qNm4IGxYUkoFfH+t E+uI/WV38S9xonhRgq2qLMWLnsRjmEq4SesIC610osPbLWoWMVQqI/xbgj25nIiVBpN+xGxyOXm3 4+tasa1qL9cVxQ9gJrP9GGZ2h6DFzlp4Zl8HM+mwiRjSYc6oBztrY4B0LAd7h4X9wmrherwjPgAf hc9w7kOaYN0l+MbGxm/dV2/vqDP4Tfwp/h6vwMGOaEJ0JgYRJr9lMwHDmSSiLrmSeUv8w9DkHQZW kv1IH7I2yzLrySFsAPN2DvuEedXL8XNPn2B/dW5IoP98JjikBowoirNFRL3bOnWfntw0ipo/MGhg G/9WywenlhiKjrTxHxjaq6zx1F1Puhx6Uh52KrMyW8BvpI2hjnT1dF6jtLxPVLGpFizui9DRr80P JQ2i9SN09goKE/Cj9XfOKVOfd5y7GAd9rIjCBFILWgV1sFtd4BDOvXQIr35mXvyWvSPbXNiujdui O7TdMmH7gy3j9nprLmzX7Rwy49I+9vnu+U4ZWtBaRv9nkwbfuBCSl/5Jvzz7R8t5T/KRTXp669J+ t+cv3q9hD+hWH/HbUTth+5qI+1tPb9HExs4IL7p0oTSsnvD2+OvDBU8NKWjV4ubS3ep9NiPUIlSD jxZbRKUHJZcf3LrX+5Bj9oXtlmuUGYPJyR0bNrgx71nCcfKp+P5M/svs8tN5ZHpQnluY0tfu1PzU kkPb2vg3XN6/fWp+QfiK3RtCV+XnRyTkF6YH6dTfKGnis/r5+e+hT3O/ClVT3X5u3pvU4ib+jf2f 9s1/DW7cXSlXKpa0sqOMmkZAXeNcOUzeIh+Uz8g3ZahiMTorCMVYVT42r/fx60/7dvBXvKqFRQ44 HX0t317zzMshvNANG7oq0j4/z6G486d5dfLzmlCJdX/6LN5HeDVoc/2a15HGvtfT57mEpDQJSWl6 NCSlWUhK85CUFiEpLY8uiWwdktImJKVtSEo775CU9iEpcdFEsY9bSEqdjiEpnUJSOl8ON3413Ldb SEr3kBT3kJQeISmWPUNSegW9NU81wER3VPF953pgw7Cx1b2BCYT5H5caBsLcIJQy6BFybxMWctOp vV/PdTYlU8pSQq4NCrlltkn3xg2ZjmzSQKiySXcZS2DIbYcZCW38M8da1jM2OyGpw5MWUDA6EKa2 KXa1DfMPAmSTBtlF4DHGL9b3m5HgN7Cb8b6ErhowbyrefEgQTC2GhfZzwvA6hNXzAq1mbPV6xDCi if/OUDiWNl6NJLYTh4kLxF0il4BiQks6ki3I7uRQcgI5l4Qw8k9rs4KMq7I2NatTdMVc3ti+F21X 5v8mSSMKNlCf3ZNeOf6WwA2mPPMqpVCfDA0DRm6oapq20fC/TVNz5i1qmkLuGl7ZIZv0lWPb8OXh /iEpk88ER4abbdLCINjmf0oyLZOadkl/kymebAA7n13N7mThKOva+gz7lH3PVrB2XBMOOnODuDHc LG4Zt5n7jzZKY9EyKdol7ceDDz+NR8uksfxJPoWvJFO+8K+NErmd0EfwFAIEmC+sFnYKR4VLQoaQ J5QiMsVJtBR6iMPESSLarYDfyxXic/GjCJKDUe5IXaXmIoqjWiFFSfslOC2lSznSO+mXVEvWyx1l GCCPkmfIo6WN8t9ygnxVhofyG/mbbKM0VDooBsVbAbRXvFZZKCP+IFMpUOCzYqU6q23UXuoI1U8N ViFS3a4eVncod9VctVjVYuCItcC6Y0OxCdhcLAzbghkV0hnsJnZWNaclQWf8/6QlXcez8bf4Y+wH 3phAS8XgQ0wj0BgpljhJpBD3CXhFfCGQ712dMJHHYEKPq8hjMKHHPajWpIk8BhN6bCaPHWhoRnel Pehx9BBqBR1F76cBMcZVhEJHBlBDZV4uRf0UmEMizBkRUEUR72FQ7wT/SxEHc2AOibjLXWRR7wTm kAhz7wTmkIj3/DMO9U5gDokw905gDomoKf7kEaEA5pAIM6EA5pAIlBHRRxqbZrJJA1OCrv62Scea +JQhaRel23eO/De5BjS6HivPlgdLnIyQMkiTT0tFco6EsBQ3pb8CiEtZrGxQlsiCkqo8UFxfK0+f 5GZaqJGRziXVPkGp/Fj2UH/KgeoKNUrdr8Jp9YCSo75Tf6m1sHIFU8FFrYPNwEZgJubwKgYPsWvq N8wGb4g7YkFXe+tgUL222HR8kc1qfC6+BLuCQyZegH/GX2LORBuiF2FFQG+cwiOIOHwVfo44gv/x NNQh0NPgRcBhIop0r3acDGl3jpTJj+6QTfjXCCTdqP5UR3I6tZiCDdQS8gT1N5lFXa30vQGtKTan vejJtDe1ht5FwzF6N3WP/kiZIXNwZ+62v9r+NI2EUxidzMANxsQh27LI8+zPgi9r4pCRcIpjHrGA hJO/lR/7L2Ui0CCQW1FJkveib3JPuVssIKGssvZcU+4/kvxvHpHkpzgkfIs4MBGmRcxP5gZ/26io 4bgwa8VZIV/4JOTy9URrAUy0gqewUpwv/CPuFOCOUT9dFTVSbclCdJdaiWCiFWaJi9kY8YbUWIIr kovUlv8hBVu3kAlpigwjJdMa4h7phNRELpXhpeSkDJNtlbayQaHkSAVCZNMa4kXZT1nY9osCepVh 3elcpa86QZ2txKFd4n1KtvpW/aGaBDQYMDejsFqETcEWqxvUxRjcx1LVL9hS7P5DFzwCy5Chg/Go vbG2OI8n4tfwaAwK8e/4W6wRgVYVb+H9cPDB1xCL8HxcJLKIl8R9HBKJeqQ10Zvs3qobEUiuIKE2 1somjBDJ85hJIMPqPnWoPpTJFplM7qR+A0u5/9kiEE/9oibRw6kIejYVxp6jQUMepfbTDyk7JoPu yngwMI5BnPAcej9zmjlA5xgPm35C/0uX066sCzOKncHCEhYFRSSww9mH7GXGKRScWIkZtiKZGsqP 5vqxyzngOJPdcZot4n5y/xMospCfwe3lQeBT+Qf8a/4r/4ZrIFA8YMuCYjvwRYyb0JA9IOwjoEoj g4AkMpiQ80DBS4wSrhklMqQLZo2sF5FEfheYihclaF8l/7ZJvxKrRTvbCLHGAqDES3cfGItAM7mr 7CGPkyFQ9pCi5P1ytHRXSpbeytBY/lcqlzAZlFHKDGWJAhuVpXKCclVJNIrqu7JDLeMVt2B7qx2U zTxC0Cp97MvqcSVfRSUAVQBAJYCsLAEf1X+w8xggz+OuqsGL1eb4JGMFAFQCuuKmEnADf4LDv3g5 /h5zIZzwiHr1lOiGEExEOHm3HoaPxw/gx/F8AsxZd87EcBK54caPwQKiqj/wMX4MfmfddSfnUAEk R8E+isRFsjB3FfWErEmXk9Dxt6M5mtpIL6VO0pBCP6Pe0I8qGfN2TCMa0AbqAmYts5A+zuyh6zGw kHzKPKT3kTP2NGY9WHcG5rJo7fQge4aNZ56ytxj48t/oqws7g6scfaE6bi7jNZTKkJgZfAXtio/h 66GQmA0cquPmMl6dB54fIZhGX9sFd+GCAMd4VMdRGW8ptBJbC3Din+tk7zfXhEwyVDgrHhbgmWhQ 0AVvLnWThkjdRZgjTRDnilniFvGg+Fi6KcJP6b3oIneq3BuYKY+RYJPcWRBklPWSLN2W3kpQINeU TRMsBJgukuGEEitLSr7ySbFUv8jWCtxV2il9lBDVBJjuVOCeKqs9VJTz0hzrhnlyxit+R0DTK6No qgRMYQWmYh9VF7wTPhAfjXtg4IwHYh+xU3gabgJM4R2GE9b4R3okMZVYSHA4DCei+3QwVnJzIQdT JSeJ/UQkuZ08TAIKPbxMmHeIelBgruQB5AzynPEJh1PU2uN/JhrCTHoMhRINN1PX6OfUdQqyqRrM D6oq0RDWMYvoE4zIZDHKgn3+/06HN0xpoIbBGZI1TakBjamPsbsZ05S6gIHPDMY5s21YmjNNqQGN qe9wF1jTlFrLQVHEqiPddoafGKWG8v9nilnA/Z5ituF78SN4P54U/hHOG19r4YWgCBqxtqgVuolD RBgvDhUmCHOFMGGLcEN8IgKCyDHji91JaiKOlmZKPkslNgU/k6BLTNZvTdX4zBFZ0YScV97dqlMf KC2VeRkS5ZliG+6B3Il7KzVSCAWqTn2Psl4WFchSUuRC+ZOCTr23SqmA5pZmrPioIqtQouqw83RL 7CVdMcnK2L4Burym9u0pdkutwAD1b/YYat8GY2OxzTig/s3Uvv3Ai7DGBKD+rSYxBb9EBnXwIU4Q YHrUUfhPPbItWZ8AikTQMNo/R+vn90iQyRJSR6HwH3fqL6oHCQwVTm2lfKghm8fjl8n3FFRQdnQT ujM9iB5Dz6LHUrCZjqOT6Ou0KSKsJgONGVemH+PDTGMWMeuZxTT82aOjFh3+7NGPspdYOMbk/d4k b8X14IZxMIkL4iK4GO4Qd467zT3n4A4LvAPfjO/Ke/Dj+EAe0HLIfv40n87n8O/4Xzz8y+mFjsKA ytyPJcJGAdByyFXhYeWQ2kZsKEINwSB6i1PEBeJacbd4XIQrYqZYIH4WTdOMXhKMkHqLwVKktF06 LKEECPhTmCJdCnPliZJZmCJdCn8KUzTNgM3KcjlJua5kK2+VH0pNFUxLAj4q2hFA0ww4qe5V7quv fg8r22GVw8oAbD62GtuJoWkGZGCX1VLMAjf72mAytmNwk6/9HIePOBDI2Db72mAytk8TJl/7FwG1 SD2JEQPIUcY7u4TcSMLfpCkkwpwRAR0oA0WQUyiUEbGbOk7BFcoUEmHOiIARtB9NUVU5bmAOiTBn RAAKidjC/D+TjM6sOSMCUEhEErsv4Bb7jP3Aqiyg7qsLN5gby83mlnMc94dNuoVD62D3w5BNGj/r PmfJ66yiHK2MLzqg3K9RUW+5xfwGfi8v8IBK+TRjKa8uNBDaC6QAXsIrPlRYI+wSjgmXBUkA1LNZ inXF1mJPcbhIizBPXCluE/8Rz4t3xBeiIkLVkz5emiOx0tFoKXBLinZLitWWFFer+Ku+zltSzqZu uXU89Xjq4dg1h645R1tsSbHckrKuwjbWbJPmXN+Savh1zT6rZjYf6fQ0/mro4RQ//ZY0y4oW7e5r Qp2juj7ZEH/d9knG2WC2mDS+FjnFZJ5B9n7sHO0nB2eabNJ5mctlTs59PCarMBVsZeu8m3Jn48dr TNa/qVaKJy63UbbLG/I8tqTUir+eezAf+aSvrmshvOvU4zdeO0c/+pAlPU6bcW7L9cOpK/O14HM2 1dpzy93DqRMePU7bs+Wmy7OUuf5v9E6GKS8NNhM/j4SSkEdzv+Qv+uq5a92xFN2iAosVRXmvH1l2 3vPkeEn9+FTtfDY/NV/V//ZJ31wr082PzcoqzYlo/664+Oa1Mrj79abVFEGXpHlyrUzz5usx36PN NQWvM5vVidYezs9+lqqx6Gxp0wnBpB20us5Wln2sRlshmNQqp96va9UfPbjVQ3PgatiztNqfzlxF N0SRNFmc1ionYdK7sG9BKoYvUI5+m7L76fESuxulKwoCNBteh+vOlO9Wpzklfwp1OJiPwRHlonJP kZUSRffdObqDpl/dzzWr8yeza5T1Ka9FfW9rsJ74OWzql48hj6e8cs8Nebzhx1pd+9m1XfVf/UJt wvL+W5k/51Ljcgu0Mj8+r6JJfKpVeiunM63CHJ641F388pLLmSZ1dKkuVhVNNXeapDfBOjtH4cfV gHo67QX77rqxHU/o6K6We61yGlPfw5ItQjrUDGmpgR/lmHhywvGVbW/YY/rXK0tLctpo0stHbl+B xcXESy/UZ907aM87rO2uxQdP6mbVQaenOo7QGOiu0wd3Gj632zyqp2HV2Jzypr+umTjQf5ZEkvzK KW90b1J01UDj+Sxlv6VLyzNXXQmPfm2fpb3AfGMMlk/aTFK0f0dq++A+L/qOj2w1ji54AAfzZzmY piSmIUlQW8KCpgj/9n5WBsF10pbUctLw7b73k1UdNLd3EKbYMNPMxNU8M9FpZ9lPdl87ZsEicpBv ylbXudPuj6z2K3B6AjlNmrjlSUjFTRL0iCYe8pN/+JLU54QFGqp11d8pJx8uqt9/zALnNXajF/Ye UI9ru1kzZoHfmnabYPrCWVzlzsIQp2i0kQRujD3dlO5CDza+NLNp99h6T1EvAZDlhyQfPKXR6qml Wle9jg1lJrf10ge32Op6Z+vXVR3eRs1gndsP8Pnp8v1K+P8f6E71Ph492X34mXns6CNzp/0ad/yk 8Z+7yn5OJ53/gLIFt9nnrHDZkqo7yDD8jO2NYWd7D6gl6TM0w8+A7w2XexPO0lJPgwl9D0PsexA/ kEcba+U3vlms5nNejFgPb+MKrqIyBOvvd9y380X3Y72ndtxXI35FlqaV8z+z7NdXX5G1IuvZrZ/b fvT73oX6ntfGYBggLrZ93eVS9WllX1TSUFr/9uxWmG9Z3Tpverx+vf5LWIz6raV4Tr39veGCyXZr fr4496+o70rXSvga+OneV1XvEbZYqePg9/GsqM+MIarrPw1ro+lAaGdnDeium51VfWG/2VlCqmbW /TZ1S7XfC3TqG8vqX9qVjqtbqqn9Rbfz481LP6TnC4+6v9WCdlL+8x7qu9QTB3LJJgPwR436+bzo ade7NlM7aELOEvyYV4RGC8G2n1o57W/68H6pJueTTv4+tlNC9S9J3m0/2ZZqYtvmFBkCfvyVF6xq nvex3j3LEK/mnVVdsr1kTzzOBuG3CRibh2f2FGt/mdG59hfLsjU9ZgwO+FSJ33a2O04a+1anggej HPO6kHhm3vpMpW+vg6e8BiTH/dje0PnAavuC+y06H8vD4guynkHBYzzTqLDU3/Bt/peC59CnIKVC rQ4uVt2Ol6QleQ2ZWrE+s+ARrJdNyN5/g36Un1QVn+RCghnZMxN7LuVJrsejrV2sXSwLMl/Z21fD D9qWfI6DC3H1HJbYa2oVZG7ve8rrbhcT0AdVRJ8J6IMs6iVlIvpMQF9KFdGX5OsNoUya53ynuIEw 95bnjWBYfLfRG7ogxwmOWNZu0XGfTa8pKQUPBlUUPDEcsFjn6Zz2/Evxxxsa6J/mOQdYU5zArFLa YgpT8NThcsKKrCTfYCaeeVeQNbznISonbCq4iVCJyt7yRHkCt/wQIliQVf9yQlRsWhTjwmY8mnIs NTVwiMMtz6lD/DL95zxh5rJBiXA5IWlsMBvPdtz30K1P/N0uYE1VIaIILQFz823qvV9w8CciitAS T5ZHiGjSHo1sp/sc9y0p2nHOc7dKG3jxjGcdX/Gf3ZPmHWIrA54Snwr3EDialLOrEZgiOk0JnVrR UYQWYndxqIiEVVilsOrReonfsZR+wwf1nh+ZdTA763t+u25Xe4fYHUu5Dt9cPybWJBBRagJKARGl Ni4LpD1VyFGZtExCUZ0oqdNfBhTVWVM+Il+U78myXILC+eoYJdgf8X1blXijBLulPFM+KKpir0JT Yzc+WB2r/hffd8qoxx6rRepP1RbDMTcM+mO31enYYuy/+L4H2DjsK1Ydb4C3x0ncC/+WgwXgqxpe jo/ZGKKOC5k8dLa3+2TdULcDeDJ+A8+svQqfgf+yJDY2UJXc3Ix0NTc9OffyidzLbHxhcmxhckz3 5i829tUrJyIHN4tf+tGz6df4kLEusbO/roydHNBUxkvwzUQdQvEtTGqVe6VN4d0W2dltVuZmvG49 wyjh0WJZvCY2qe1p4lbrHKKY2EQ6koZEsjv5gLzYZjK5tPPsleSYbWRH6jz5VRk1XJlumPusd/aJ jcv7xkeuHKSXW9SkjlQCayd8qfHDIJe81WJW57F69JkfSmdRFUq/wvQThenxhRWd9Jfjcy/H5ibX jxgZs25k9oloT7bfigFhEX0/d21Gp1bavXl9v8IPGgKKKctJGU089Wgb0TGAmTc7Pc+gav3ne2+e DIvptT5bJtxo0Yt5wvgxwQxETq7Naph6zKXmN5mp7HtmmbHDDFUykgvTDwUByh+I8fHsOliPKAqr zfSJyW4nxtUPa904bGG8t0uzsNihzm3CymM6hPWz29jPOrKRe/cIN1hqiGjM9mKL89jlnAWX5jq7 n3Jiecw/KwtPbCyMTwpXc+Mjc2Mjs2NfLBXDMmJC0lamx2hC+oUtGR0+fovn1EF99QxnHc5t5eLD kj+GeW6Y2Xf6gJrBnl7RhYVpYWRY421DN+b5D/3GbeMb8vvDhyrJU77TA3YpXlvHRbtbJyw3hI9x mNZdj+TqejbkyhZV/XrC22Xj0NaF6ZdPtAzxivN2t5uXnGd4ti0tbKO7dddjkTP4mgdiawvNhW7C kLjLy+Nrzls6Ljekn09cPYWKOhA02d0nbnJcS33Biy2KGNY17OqBjPjIFBsRmxYruIspgnXCUu/y rzn7MyJnF8aEXMk7mJ2Ru9Q770Dsg40hA7dW7LcLO5hxYv8STyUxqJNLD31swwuXF7URe4npYvbS cV/HxFW8S9gYkrx0svE/JDLvQPrsoTMTkiePdS89JIbFDOXO5iaGad31l7s1/eeMvf9p0UvKEXPf JUSO+5pekpgbs3fpt4TCyydmJuTFRRb2cync6pm/P37jUtcZhvNhbcKizijpe043v1rQ6W48Woko zciNnNzta3ryitT4jd8SRkddss3cmrJxqEthG/3lxBM214ri9qd6u0Sd+aorzA5NWDeD9RzxWJLy paufpI1WWeEpW42/TiztPTn06uyYoa1x4zeHp1tf/ZlwKqyZ8S+rE6xCaIJs2G336MFeu71OK+Qg eYXNv4UbZ3/NuByeEluWUJaQEeJN6G3yYn96u8ckhnhnLB2X0jvPIuVC7r55vTNlg5LNX1YyLhfG z9uZcMKuUA5XrJV7mTGN87crK/5p2Co0ISx+XOvOj+XXK/1XXGryNTf7a3Z6kGJto0YrB16FHFMu KC2D/gkLaxw2UH1X9F3ZqtZX2/1Ib4fnpX9OT5ZKTygJkxNSHseJUfmlR8IbhJ1U05M18RebxtkV qzh2SG33bZhlo9kN8wo3To7KtMhclALvE/wSPOI84g6GHQxzDuuBaTJ02t12y7DDGIHdxV6d0Vyu noXmPauqhyei0brdeWwU/sImxPlEVmYEvtY+LXlX7fi42v54kpMpIDnN7pbDbzMzCR+K/8BrEr/N BDQL+oCbZkGQUmkhzSbKCLQD0YcEUwzNT8I8CwJTDM2GbsbulhpAjaJgBrWEQjnBCdRV6iH1hoJv lA2NcoIRlYzo5zyEPx+nr9C3ahslIYb2C9Mqp4JpdoONmjCmh6VRd+rS7DSO7ow/E8KsMn6hAXOc gStMJlPAfGas2F1MG+PVhhGsHxvMRrJ1K3HIuywgiK6KoZvAQeXP1eEOcmc4NBt4z4EpfcLM0C3j AUF0SXwVFPd7NmBm6GIFQBDdfeGV8EWwFuuL7UToI3qKAeJ8cbW4UzwqXhIhQ8wTS0ULyUlqJfWQ fmcLR0gx0iHpnHRbei7BRwlkB9nsEf/OFt4vn5bT5Rz5nfxLhlqKXumoDPjtEW9UAHFxV5WHyhvl m2KjNlShg2pQvdUp6gJ1rbpbPa7CFRXlBn9WUW4wijkClHP0Z8wRoJwjBEOaIyYBUbFb8IO4mbCA CtxMWIwhEBQLiIpNIq4T2cRb4gdRk4TGpGslYWHGJQH9oKP7JMIlrSkUMAvmH4JkxiUhg8qjTGPg VjQKmIVJdBAdQcfQKFT0Nv2cho80MA5MMwbNBdBYAFB+CJoLpDM5zDvmFwMoP6QjO4BFcwE0FgA0 F0AB0m/Yb6wN15CDDpyB8+amcAu4tdxu7jgHV7hMDgVIW/HOPPIDARmCaG91O3+Yv8Df5SGXL+a1 gqPQQuguID8QTIbgQeGMcFN4KvwXKtFZHCSOEStDJTaLcWKSeF3MFk05woCChPtJKEd4kbReipXg pJQi3ZdeSV8kaxnlCAMKEg6QUY7wTvmofEmGDDlPLpUtFCelldJDGabAJCVIiVBilEPKOeW28lyB jwqoDmoztavqoY5TA1UwzfpOq+mqadQHtTA91hEbgKF8kSXYRgxMs76H2BvMNOqDDrgB98an4Avw tfhu/DjucwVPqx2YZqdNs7NKs7N1zXHwdU6ze2efVk+295Ht0+pochyN18wize61ZZrdujRnJxcd +FLI7qjRLM3eQDQ9PMCFbFG9mVOOQ+gzO785+rQ6lhV0E02ok77jzNYbchwNxBfS+Fy0WllMhhCr Wjq5bCXi9aVspdQpQ3TlVcK+2waX+i5viJ6uP4npBODkBpfGLu6kNzGRvEVccX3tkWZXu26zx24a 8Jk4oG4zHXRNs1O/kX1VJxfr4RZtu3Sp5TKjWivLhjkOPu/su3Sxat2reZp98p405+/RPewMOzr5 edlsHODkuqOzcqh53qHmZ2XSU9c6RXeso8XQRPc85y6WGa2f9a2fY/9JOz/J7bOb3pxqVzpQN19y OQneZ/xq4J7ppQPh3c/h90sHauqMrhuYKqCfvuLsalfbRevzzM2/n8VEb93kUZaP7cNquaBuTEeg rG33Vn5h0eT/1957ADSx7G3js0kogseEYi8Jggq2gAWwbgBBsQUQVGyhI0cxdAtqgjQ9loAgoKIB FFQUAwp2DCigHktAETsLNhCObgAVFWX/OwkR1HPOe+/93v+93/u9ji67O1tmdjM75TfP8/wOoAXo rVFLnR6OKXHU6gmWOiEWPshWttxlnx1A+HrobfYC/t6ZcAZkGMo5xv8w/wpfj78GLdUbQxM0HHeh 39cNynfUAuBde2AQsjoYCis2Oy0fMt2mtKeDe5A74RvkHrA8eCQLc3b3WOl9MsDBj/AMCQ3yDgZ+ DmedIcugtE9iaR/Du87tpf0Yw8CYoAFXeIEDrhSjuU7CmnHd9O/rP5tMaUHJ8S/35Hz9MXMuOBpV ru5eOqByXmnvq4908h0D7XvkOywK9Akh8vVWAj0O5z62Wy3SaWMiZ4MeF7JyJXZIQOiXTV/KV/o6 LIPahKsnhEvzHDdTR723fh16XOgWuj7AjfZcCmm5ob+Yx2uvmpREFeogAe2D7+u4qk9LGZci1F0Q 35sZ7DwxftxOKhqvvj4Bsd8Jpu3cnniN+4h7dtfSeZYRvq5VT5LU9k0wHvQs6rk6seKuNtDlcAZg I5e9OxeRyKmvJb443gD6HE4ioZHDaFinxUG+AJOlp4tnMcwXfqnk8Fh/UD4Kl3984Or4bs5Hyzu3 LrbtrerOY5UwqnL1RJWujgciqBab3G2eJFk/TlZPyXicfEhyNhN5ksR6cSgv549kXHI2LdppuVYU 0cHMjaoavgkgS8OoUgcFNTffYYOLOFbkdFxUtHVkvstVngtiPmL4AsFxZ4QySBA7SlrQI4GScl/6 yJNyz6xnpyRygkieKdoovSZ6tM3aJnwZkTW5L4ezz/zSlQGnz38UaYsPByw+D7q2Sh3I7hLxEmez YvK1VL0Xl+QSRXfNWHaT+3AEnKdX+4mY0kQOUZJbkqo+pR8ZtfCgX/bcMyMlBGCd9PXSmu/15dF0 Ccuj97UiRpvaM3cJ611149XW6kNPwiWsCUOyNOQWwLU25UBVVVLd3Xua5w8m40feIFVV1XXZr1/c A/X4ECmcN1ks9ZfmyeKlQKm900sGp02uE3f9kLmlDAZywRHy9u0pj3Q4C36Jajvs/Uhnom2Ro3Hl auFSx9K+RY7Va6+HPtIJtB//x7OoUl3LYUZpX8xm33nATOTYUwT7OPPex14x9W5FmqkSThu1puis nWlEGfck1sD9bJeYgkMmuzYPfHXYM5THEhz6cLjCZGoDBxewzvbTOHszwdXRYPQyJwbiO4+BdNNW X9VEGxBcpUs5o2uou3re1V6GuouqdJHbvagpE4wvnCUef/Fq7kEBFNOIrxC7NrV8h8EMCqWCr+M4 cUPfoOAbeibAZE38YCTZkLrfZOm83Ku9OJ6Gujd0kVSn5a84Bka0XmZsMLnt6rmZHC82J4w9Itxw nXmvhFVG+kw+WjSVYRJQgt3utdL5di+12IbM5dPvzhtjtCnwGJpLdpXsds3rOa0GNQmoUbcHuNVx 56WOU8tPOhU5at4JuRO4y2njY0Gb9x7TXQvaTQI6VBr50A6yawmYet4htp3odmhZc9hShyul7cAe 7HKxp6scdozkAZVchEotAnzrsGM2Xx/KRXwJdJtSMu2K1dzAmQesNFw5mtf01wcAm3KOf/MEK1Fg YGAawi0HgeVUoDc+iix/ZtsrUpF4g98cdPqXhwwE6f7TTAIN0lw5zhVaEanFewOHunKGDd9JM2l+ O8KVM5Lcqcg+o8N25Zi6csxcOWN+c+WMdeWMu4u4cmwQC+en5HqCK2eiK2eSKweZvHm5atJf4R1C 1bqrGvfjgVucAD9KGLKeAwnIm4VBYO3WOZmiua4zegG123ZSK631tHdFrjZ2rvYqZF0BK6cVIuui hJ3IujyRodh1pq7QNBdxCIw2E3uJXa3mrC3jzpdv8LYOUFhVNgshsm5z3B7xMfJgPzeh6eJ3XHGN eLvatMAN4iH2m4XemnGieP8U8SiJ+lsPChK4xUziJZFaJbiD9aVbvVcLOh3OgK5yJxBLBDrBREos EadT7iTwOrKfU0vdUB5YfGf7XvcOAvIejwMySED+nKiUwujzdtUX6IcGHJqxQnse5o2txrZge7Ec DEDvE8+wd5gG3hcfgU/B67m4O16fyvkcB5F1nxNtkF2Ie32cZ5/mVESFrDvtBnJqr3O7SqSATi81 SokUcImYceTW9zKGAfTvZAyfEN/JGKocGH2VMfxC/07GUOXA6KuM4UD2DzKGT9FvZAynoD/IGFJ4 38kY8rnfyRgO5n8nYyjifSdjaC34TsZQwv9BxhA6KOkiYygT/CBjqCRVfpUxxEW638sYKkmVXWQM DSTfEWkhgv4bGUMr6XdE2rcyJZEWTMbmYm5YEOYug9rJ5zCgZNdBct13brAiMKUbLKB0gwXJdXbE QuJXAqwndhBpxAmihDiJQ3gR6E6H7DpIrvOhr6H/Rgep9OP0Ivpt+nP6HUKT2Y8JRjKnMh2YHswQ ZgwzhZnNBBeYt5g1zGYmjd1CN2ZPZIPZ7GXsAPYmdiI7i32G/TsbPGa/YQNUFzVC9dgz0EUogOii ODQDzUfL0PtoAwo+oz24yrHjWHQ+N9BWiayL4my27kDWBXLhLAJwt73DfTqr+KtDA6CClvK4SmQp eMK7wUX4cu4QvpKOA1R8nJ08JR3H9Au/2fW9wwBBVJRpIDMYaPFxnptAjx8lUPLpwA3Beb5qANlT MEwAJghGkQNIn68DSPCH6JFAOYAcKdps7RIJlu22EQnEcb/tF8eId4qqxEA5imwTKQeRgyTARewr 3iM5I94nLpcUi99LgHIUOVU6SgIHkX4SUCQ5LHWMLZFuPVYubZWqXwNvJGu3RknhfOVMqUAWLwMH ZDulpbJT0leyhwrhK8CQDZZNwfywcOxXmRjLw8Bl7ITsBaaOdcP745BWCxzx5zkPc25gZLuOp2A3 cYDhymHIEMKCgOwc4E8ohyGniWvEGfw1AQhCh742Zg1BxZQDSaAaSTpjNfRmei2hoGD1pkNFLOVA EqhGktfpgK3LRpgdFCxIA8KYT+Gkdgl7U7uM/ZGtjb5nmqEsdgcFazl7L7qNLYWDhGfodfQh2pc7 gjsAdeRy0A4K1iY0njiGYlxzCDOZwLVh6vJ+2zyVN527ngdWcDvhCBN5WnzQxh3N9+IN4dvwFvJ9 eal8sJWnhCPc4a3hi472EABLQSzhiL3nuwpCBJH8MwLwu+As/40A+qZQ+tsEC0X2gmhBnGi9KF5w QBAvAg2iB4Ie4gRRg/ME8R7RSx6wI3/qX0U2Co2zR+IjIgCF7YF4vARCFmrFi8QA+iqJE38UK12V NIjBNQn0VeIidTjMlUBVWjBCxPktRXJPWiFS4k/A/oujZPNlttJ1Mj9puDRXBsQKH1XvZXVSpUsD UCjTx8Iwb9keLFKWQpRjoK/0kuwc9ofMGH+JQc4HCMKj8CQ8GlNiCeXkj401YdDwPYOYgEOXBmAn cZCALg28iT+Iu/jod2A0UYd7td+UeTD59EXELjqAsDIVqkyP+Q2wWMQU0guYAMLKIKqMzm6nj2X7 MsGwz5sH2DERwp49jjjPPisBnf5TIaaMhwJIsYZOev3Qw+xH6A02qGZ3xZQJuA1RVkyEQ/k8vsMY RZfuR42370EN3wNftHJ2I9kIQDcWbrwgHojiuXEhROEI9zn3JhfwgTmPyuvJG8brw1ciFMBBfgIP QhSu8e7ynvOG7yA/8QEERJBmMZXUOwA1Skv4HwWwCYAtAIBNwAJFE6CuoN6BZ6J3oueCvmJN0RRx GNkCANgEzBUrmwBIvQNUSU8JTTxBMlq8Z7cZ/0gq+E2yJ+nXI17iYPF5ss/eRSDDVOIthZYkshjs kKRKVJakldKvAhkO0mjZOukhGTgrWyC+JyWW7ZM1SY2wnrKvDAm+7CCWICvDwH2sRdaOKZHF0/Dx GFDyatNwEVaCn8TMcCCSNuN/YGelwv7mhBvhiINOLcpCvJmoxUEPAk6wmhCTiDmEkB5IgEgCtuOq ZtyQrwCLC5m98BniAKYZBIsfoKv8EMJm3IAJTjN92AuYfsxw5nG2IwSLX2Yq/RCORlE2B7Vig9Jh j6UuPo/Y9dLtbBlaxAYt6EI+/MCncKEAqQMKorkQUP5KYUTEuTUo0OPRuBN4s3hLeZAyG8AFmbzZ bCU+9Cb3KRdAzUojnit/BX8DHwqQxvFAKT+fB+URoCsLCCgHzxUMva2CfQKJoFiQywcvBK0CJ4FS wpArWk4nP/Fn7N2io6JCkUxUK2qBjol6i9VFKrcVbiJgKo4SqYuvi5+IcbHC1RhFbCFhidVxpduK Q2LgLTly0Y5syVUNOVC25Ask5yTKHh5QKqVoyVQNOVC15OukQmk5WYWD67K0kUodFKUMCojAAmSQ BJ8le4S9VRoMDXFdDOqgKGVQQDoe1+GJSEP9bC21FbTjWp/74ha4yhsRUIrNnsDriI+ENv0TDroT w+gQlaTyRgSk9Ar6M/ptQoPZlzmC2Y8OEN19Jtw+u0etEmxXwIUB1EiAcGFt9ic6m03+lPPYKoXi BWyILANdoWUQWQY6oWXZbIgsA12hZRBZ5pzATeYwb5lSr41nHbVCnKNRqGMKZUzhtws6f/Wl3ATe aR64xotAremNvFl0wBvPn84Hnb/6SX4G7x4fvOLf5xE8bcWvDjUxwFqB6le/I7jEbxWAbqL+ogoM FbVhvd4MIrtvAH68KhXqWkEvMYD9NxMR7L7Buc8sMYD9N5UKNSI2lwDYfzOSrBdXSjcfXykplQCV /ymls5IxEuArXSvdKoU4NAhDeyEFKv9TqMxR5ilzkoJY2W7ZUdlKmXuPYPFdKQ0DSk+SKhnDQNlX GcPrsicyXGaEA5W+vlLGMB77Ucawax9dIWN4Gf9OxlDpSbKLjOEz4jsZw07/Ux0yhlTmDzKGKv9T HTKGhuwfZAyhJ8lvZAxd0O9kDJUD0y4yhqHc72QMlQPTLjKGu3g/yBguEnwjY1jA/0HGcJ1I6Ujy kqhSBF6K7gq0xJ2OJIHSk6TSkeRT8VsxgJ4kNcSdjiT/fJp6mKRzmhqckkJf9n9I26Ww8zVOBuxk C2XTpetlO2RpshOyEhmoktXLPsm6Y4MwU8wac8aAD7YG85WlYsexIuw29hwDEAraT4H/hRrTIWQF jafg2fhu7BZegzfjNLKHRRgTE4nZhFJjOpEAWcQZ4nfi7NNa4gc90kg61CPtgqzLVpiDGwBE1hU2 NNAHMqmDGPqDyBodQP7PKgZgxjOVU9IANuUbyabcgA09nSxgAz/2Z6ZSmfYy+y67jg1gn20gykat UOhebjX6VRKjAn2GvkM1uKCzSg/mRnOTudlHuCuzOZRsjno2x1S90NqlTzZHZpVtX2JVYnU4X+Oi jakOLZujls3Z0j5kgApZJ5+WbcXRn2fiaDT/NK3Xw0Lr0CKOGyvbVq196DQnJNSUMdd1a+G0Ia63 ZM0COUrWFgvlaM1C3q8LTHXW8H5zUCLrtjhAVv77BQGOhBUYwmO51fBmk8UrwJHqOIjvYMGz5h/n HXCbkc3pXTit+oI7RNZ9ngaRdRtGYt6mOq/9X3Fx2+VnsqcVWe11pwBnmZWmQ/bsIqsQF9x2b/YM gxbOqrU+DAPOek+hVmjIPNC01SUmrDZutUNet6NF1DgPWpJfzRcXtXEnXUsC+xVaUUKS3R+4dyLr HoVSQ/IdXwVBZJ2m8OqjUPB89R7Dm5L4lzOePQpF2ldHzteKuP/0i4PRKB1KkXtVixVCmx2tNStG i/a73Sbq7Bi1+TH8GIisG7Swr/48A5dGe6eI89bCFtsRwbes4ReiwevrKKIMWpgb9mvKms2CYeId /EtrvE4sKglkYEFJHh7IAW+EemvDCcHGpJvB23decJdZAVMd2Fdq2tliRexGBFdcWFCH5KHIUZDN +RjmFxz4pimhG4djMO8jynFpb9pZt9Xl1VaX+xlhtcIOwdi1E9QDcchAfidgMXSe9s8rhrrPRNjt YRv6zM21jtgUT6XIrBwiF54YRV12Si1q0MKB+s5Z20d1x4KEs4Pd2s9FKzDkNIHSCeV6vrM4jQ98 eDqylLShsmbeTNll3qpbSP088U6ZNv+UbPCV27KpMs5SviZmIIMOG7ff7rH06kfdqQeybbkB3qtm 0t29vLUe7NwTUcKvJ4di/O6S7gLTLFMduwirg6l55UYDdV8VWc3f0MM35II+cSHkCjGIw3HDsGEC j30bAnboYQo3PKuvcDZ8SnpavvqKg9XpP5bhUILyLLKZOupxZEJqrluoVlT5V1u3rFb7eOVLqlDn YN7gQiv1cU/qb9YL88s57UWet58Ov/mUer9WvWcL8vTpk6emQ99C29C63VTKngiPs1Vh79X2aV0d 5BtieYgIMvclBnI4B7FAYdPvaRsCmMXEpgd6gMXh9EfnyLp/2q0Vn2ZUiFo0Lo673vJHltlmlDkr ExXg1hHVG9AHWWJc1PPj+y9r0NaFm/hZDdYxuzdtdthkrhHaqm4XaUG374Fwwt5v1Jj5S3TrZroT iF569Ysh6AI7DL2zvIrakKuAHWK556It686dmI4vdG3EClqkLojG9eEeGCWkkrIeW3hzYc9Dz/bW BzIpLy/PzeZUB1aJ67BXvYqsWK0De0vGZ5plNuYZTz20KZ6Tbd+Yp373SzRWsGjteN+1BMuIw7FZ +8uLlvHrRWnZNu9RTrNUOVO0uvTlpt1DXuRLoGmiQTIxfYAUYuKenNGnswQQE7ey5aaBlGBpf150 hyJ92V5DsFp1Hk8PYFpbR4yevqk60CGyOrDbDopDdWBq3pWWsDGUhWOoy8wr1Jabbx0TNgYJMqem PNW6esGVueJhijEFUNoLDR5PVCDiCHTwNAUiTji0PPxObELtzeQ99cUVd4IcVo6dnLG1suI2YmFD nWrvtvDkcnN08G9jV45BUpdejXhVPDGbVYG6Ddg348A6ThHkUGx5tFicgInuQpY1nVfNii33gr9r fpB5o1ps7tXp4LLCcXuCwzF0J+/g/IzIAucVNWhsec0rBm4VUrl5U3x1TfWpAVMPvHDM2DS0aNHR 8ZkZERMdMmJjy7uIUb6uzdgGpr7MhXi42xfTYzbF+b7Ym54R/YoZxQdJ/MP8c/wb/Gq+nE8R6AvA UIGlYKZgiUDFsOGcElz9ypT9oiNk4TU+HnvT8bzcPA3bgyvNrAfKwbOavSceOB6uoE8xAiOiO1m0 kH0DoGa/ikULSbSgK4sWsm+WaEp0Mp0A3ygtZHIa9KpjnBYE1o7wP2grGbG815vclr5aJ0qKPo3w tBvhp4DBBfcpdeFoQRicUVoXrzqOZB96xK+6LIfGPJ3MwxUSmnSE+5zgKu574H0emJ3MnQvcjdMg Cs44F3JyR7j3YznEtanhEqF01BAvab7hWXtd4zTvw2nsPH9L6U3p7ibActA5UiGlyXjurnc9vT1F YLJsrsxN1kkFAN9y1Gdi4FuO+kPMWMlRRzyE1IFyndrPwTmTIpQO1ydP3GSDQwycmiwUB7E4Y3iP gdD0pmP8Yek7HED1uhEEZIEoSSBAyQKB1B9IAhnx60C3ySzrOXap+c90xhs4LTQb6UZJPcmYzCpZ IARzkBV8pU1OaZIDt+l+7ne7+B4EzsxH9K6+B8Fz5gpmV9+DIITdxuzqexDQ0K3srr4HQTf0W9+D YDdqzu3qexCYcL/1PQjOclViVtD3oAXvczHv+RIFBm4R0YGBm4mYKakEbM/nvCs8L35742IlBq6a 6MDAZUqVGLhk722r0tzwnKgsr8xwiWdra2bwGffUFa2pxZ7FnlAc7zU/hI/XZeusqR64rm5EWNX4 ddXFt6atvcL3EkCtncxUoBOuJzBeayHgkkU8mCzirGTBGFG/dZcED0NWQOGVkfGiPqJWvGA7Xiac uN2iKmfb44jMqKcxU8MYK0SaYkguzMkXnds2RQSMw66FnHGDpf0IWdrn7hR0YOB2ugk6MHC1Cb8k 1idU5cjjqjc9iaqN4MyLiRSzyCHiWTEa3+oqAbwMrthr/6ggiRv0a1Ms0So/WobGEe7pd3a+PnBf DOoSm1KHhu2VWEoVUtkHQJDUTbJG0j90mLREockxlV+gwMCpFb2WCAlJS6IkLMsN9kO9b+VIIAZu bZHJhqLM804GkUWpM/tsKUps316kxMBN6JF8yQyEiy9tkO6VorIPj8nxiJEQYuASL6rfUWDgdCqq iQ4MXPibQeW3EoON7pQlBoP0ogfJNXub4kpi0tyg6hOtWUYrL5hTLrl+NaIsqiJOPbeirs6ofH/R hrtHbhT9FrAAe4uFY0jFTLxgoWTMgQd47p2zFRM06TdrUtICs5LdIMX9lTR44NOvGLiylsKcmJO5 NU4TGAWWaNzEF0bl2yZo7mq8gl1a8SxVacw6XFOYeXFFngIDd7IG7368nFLkOeFkzaWaGLfJTwPx QeW7igxabmVG+RE1q7La8BSCSfjKw51aN1o0KzFw6FtMgYFDW1K3+QUfvLO0mYE9v5XzIM4RZxTt dN/tlhr+qXALMXkvMYQOMXCna+wB0YGBi3qDtpStmHlVXuA5YaTj+0HliTPftFYzylPckMIkT3X1 gHQ9ei6dbFoBDjFwDkh1YvitBaCuMOeqHJVHWQcb1DXHcZozt4VfEfbpU76lCH+Pl316G0XZ2b0p M4OeT3ccVB3l2dq/rOAJkrltAThVvpJR3QwgBm6LW2HOfT/qDDmi42SAv2+tQ6oq5fWHq+NyeisR HrXbvHVqQDP5Lyfcc38lBWLgNjKaQRnHl7II6JZHkomFnCAq5WBq3EeHcYafHT7Pf8KUMf1GoEoM XA1wBrfI/8FOO9z8WKlOGya0NCkwcKmsGk/QdzQoTmVMjcOrCrkQA0crbpXn2LE1a9i+7JGsxA0G 79hP6mILetDLMxeZJPSeOv5pev8gVImBk7H9UJqcTRkXrIX2RWNy1cs3FHEOovYTFqLNUHdrYtlG Vq8h08oKRlvnzAWX5ObymkE1HGvNinVFv3DLCjKF/YIJBy5Z36hxt1oetV3RI5xVt80TZ3gy7oFZ QCiVH6o5VEMtp5avLtrNvSX0sP7o8IirwdvBHcGzUS8Ey3WgKeG5fU2TgBfP68PbXMCbvOj2whxT Vi2vzrHg6Yd5mcS8Ip6ODxOAFoAYuRl7dDiH0eEf+WpcAtDOUEqOSSE/HPLQASSi/853FqjMDADa GRYJWgXQzDBKhIqAo8hT1LC6U0kffKukv0AMoAMc6N9c6d68TiyH/s0HSow94eAtdiXwNlIMRo3c tA4ASss6L0myhGrkpg9SOmyGRm7rJNCsCFTCux8kW6RQdxdIpRXSZ9IwqQomDboK70LdXYDJmmQq 4d2vqosq4V2ouwsITAdXCe+uUKkuqoR3FbxzqNymEt6Furugq/AuFG4DULnNs0N4V+HZvqvwLpTr A7wOL8gqJ8igq/AuRPEDCOPviuIHXYV3IYofQBj/ZVQpvAth+qCr8C6ELAMovNsVpg+6Cu9C3V2g FN79Gy93QMU+7uL2ulN4F+ruAqXwLvRwtVbU4eGqU3gX6u4CJftYSVmFjFXQVXgX6u4CpfDuLgVl FTJWgZKy2gm9ACoOKqSsQugFUFJWbWWQshou2y4DYpnK7dFHGXR7BNiYFfa308MATg8H49E4nBqC 5CqgYFcRPYlhxARiFrGUAHwigkggMgk4NfSI6JgaMqSPp0MW6gr6BjoQ0dPpJ+ml9Hv0V/Q2+i9M wGSaMW2YLkxf5lrmVuY+JpAwoStUpSfU/uxRbICyHdme7FB2LFtJPQYydud8wSR0Dgp4aCAaie5C D6Fn0evoExTgKMJVzRcs5vpzgYAbzz3ALeBe4T7gNnK/cEEnO1np5gSIeXk8lYd6bf5AfraR50oj N4qRm7qRm6m6hYdLHyM3e3cjv6nuU92zjbwRCy/yM6MZuak1G7ltMVoe6g7tEsMQ+m+/ehm5c3Z4 HnCv3O+z3KuXhUfoRDc3lrmRt1r7RQ8kNNTtqt/urRZeYv57lKwulsuL0dv8576h7s18WkBzP1SJ gYP6fyBgdYP7WndbwZ6ARYIy/kaBsMF9g3uKII9/QWAsGCicQabc38LLPAJaJY6HeVGAuZEbsUBE OIS6+2735Pt5L49ZbmThoeZs797daOV2TyPfie7+7i/2Gi3v4e/WElh8yohgmqYN5nAu+BOHtidf 0OJwNqAKdwo1aqu/RH0pV1vdCdv61Z0cyVq4n6eE5grJsayukCZ4rhzLWsY/0Z4CcVuvA9oHn7Bw V7dLMU8R6rrG9z6sHTk53nwn1SpefWMC4jBrp91OUaLWJhpjS83qUBO1UWul8Ywgd8rECH/3FQjt wU4qlqS2L2W5/aAXUUYerGahkUdRm5DY5G8/1Y1lyCGUfmbParE5COhn4RGiEzEvgmBxmxwPzFv/ /MbqgDEZSMjowCkZi3Pvqi1Je2afSbE8JD/G20Xps/+FfSbV6RA69uLT3+4Bj/TMCAPqhE3PdooP VSer78moTqYdkZzPRLCkukMnc94kN0kazqdFNy433OFJDnZRBaQLXe75PmwKdY0b9EDr72buLvR3 T4+d4ZErurTV3zuC7zIE2bViuFzg60XREr30ZxUXtO94KA0JNo8YEORlLQLOIqdL/u5u18JFVqG7 tkx0p/WMOBKlabRyovsbbz/v+1tef7x0ljVA/6CQxlnCTrt7CeW0O12qUwtd+kot9D7qYOQl/NX9 chE1RywVj73rkzNBHVNouG2QsBgvdNZXFNtnWhxuHyrAE6TXMHsJa8Pt6GGS8idVSWOLn+2WVqmV RqUsH7jDa9uFsBMUClU42zXerQD2jbdJQ/gghp/CvyqF2IV6aRD+sYgVi68UHMWnCGRv6vMFLTiI FvQmIuvGEmL8nsCP2IQbbyfExJvmHhfre2ZFGnkT0/ieChk3erC/78qJl17cgfAu0NUvyzVZyQdt hpH+QaAaKBUpSdNwoMRIKziJPXR9fexlKsYaoH0lMI/xuCrpRcWz3cW9gBLn3AlzBstxJbg1Fw9k V+ITCHCIrUVcZ0cQODtTAXMGFugB9jHCH41hx6Nf2Ordv+j10nbyDg5dGYKODgrxAXMp6eZ0MIPu bD/R3WPOOvqN6N3Rcw0Q46zYB/FGfnMN/MYvG2I95NP9LePVj4BhHM5mu5zANe8eDlmw5vq4lwtD i1+oWqZnl8LIXkbskzO/MlnmPExhcYhissKlbb6Hq9tf7WeyBImHw+8abLk129Rg9MOd462eJYy3 6taaMd7KaXUeMs661oly2Yl6bX6F2p35751qnZAn86kpT/2NL5SityYuWANtDsaO39scpm0xrR5L QdFN6Yen9LUZnB50b56Y8mZcqgeS4UktOuL3IOHO/Gzrd/PuOSGplCLjVyN/7ekwmX9jsuYalguY ng8N5q8nPp9oNR+ZynZGmOsE9g4NpjfFT+bXJz6Zf1QtVs9yuqfhmDy9Y+7oFwE9xtCQFVWDNpj+ UuPMwa0mGz2Inx5nNZQxtviIcGyx4eChhdo3og0NPmQGGA5rMFVAz6RKCp7hSDBwqj2rneg2fnTA kGwrYGdgbdCOGA5x5ippYqCTJ6akiQGlK4FOTwIOSlcC4ZIv69N0ZFEz2hEnAz0DjcO0e45ahiwd 7ZD8aRHdEPrhaUmZdyUAOhno6mMAQCcDSh8DsKvzTgo6KWGwq7O+9J4T4IuKQsTAFgxclVAUBNYm bc+VGRof6eXIWmav1XNckaGhXbuhiYp5V+TmPwwBNqKiTubdBRltNGY4XHdrwFyD9aWTMT72ydBg zqShCgdI54FZL9Zc4J5QBG0OCeWw62Ro0G9rQFyU7SKsEUux9RpstOOBvW5Ckbde+Z4K/wOYJT7G mkpB1l+bjPNxO4ML429Fg1uSH6RSlWCMv5RKdYBSqX70cDpyREg9kfZhPVJ9Myu9g3eXmXGcDm0O ATiAAI0Jwwkd6P1iApMzGHhuV870qCZ6gGqmR+WvKmQVu+Wh2yGO9Rw7jecz5lqlWV12HPnBSeMZ 4xAHKdHmnDGoEqmUJACc75uCVkZwFDMBW1CgnAmw65gJ6MsFypmAKlQ5E3CEC5Ti2D5c5RS9Qg8Z TvO95ybwMnmnedd44BHvNS+Ep8M35MNpPlc+gP1vGl/Z/b5Hdr8B7H8n8mH3G07zrRUA2P82Ukzz vVB0v+X9RRmCkSLuNmh1III7mHcanjPN/EWwN74nkiv6MEYkE83fqrA6EGUdzDuisIN5d46ya5t2 HJ4TpZOY+SS8X0JrZrBhfOoKjdZUz5EJR0TnRQvFmOgoXrc+pXrznrqkpCr1tD3VtyS7x4hlCqV1 ambq+r0bxQm7M8TnxAskwmpxuIQi2SeJ3jNKYntwRS50r/CZ7OW14qxC/LFZbmZOVc626XmZUZw5 p44kVUmCpRAqmMOUgiEXDksSksYdNIyDXjX0ZZ/3Ss+eUFodvtR9ESqtDuMLZl9OnHe5Kmdxkf0J MCN/dt7FrDfSrTLo9DA7D2ktkf1+6ZxUdjFlZ7+4DTJjkWwkNulRWXYucaN46iWw8I6NzKlkyY34 JHXsQAeS7M4T7LrsmSxm106MjX9Q2LnuKqwOIKAGOkReVtIvSyfOAB+LXyx/o2DePa8xeVNfk+lk 8KYmdWafT29rEj/UKKwOk6ImUJrNQLhWcz1Wo45DVuUtXJS2whpvy0kMbFFYHdrXN3Uw786nhm+R 30oMFrWUURODu8unXZl1fUkR+5R2HEPZDeXLC87I+7WNzetjlj+5qD9SVyeSd5PXvyvS/zzi3CWC R39JrGyaiV8ruCzr0Q3vTzVCJmgK172edfVxGSUOjkWc8eD2zZQOq4PMpKwwp+lOf4bThGeMguzc TDWRfNsETU3Qfwy9ipoKJaQRph6jrTCzqkJhdRgziIH3Bf41nhMGMUYxSpqEhyj4FjmQb9O7lRkR Vcl8UObDpLJ/Y95mhLc5tWboKqwOG4Kz9RVWh2y9Q6nbgunUq7qMmfq3bHOmFeHhNV8iaHGpL/s1 Fb5lqrPj2NDq8HkwI5+ltDqcT43K1itbMXMsq+Ck54QLvbbIE2e6DqumhcupcYVIQuDQR8Ub2f3R CxnsfBa0Opw3qJ6UGH6JVVeYM5aVzZBFWRvULS06qpu5zT18TG6k/G3NoqF4mbdRCT74y8jMX1Am emFqdRTfs7WsYIZZ5rZLrDUG4J7pUlNodXgr9CzMqTQ8yVhp5mSwaKhGa10Vypp3376oz/TdaBT0 TLit3GwWefJS05zwcE90DLQ6vDKlLDUtuz24lLVB/oZMDLuNECjrSK6XUGzjK/SNmcH9PJFbOUtp dbCcZSplSVm3gp0+Cis5ralOE5YZQKtDxCINh1usqLkrajU4R3JxtErBvOPXerAMcvK4s3i3ucmc xHorN57ljN7Nd9fJMxeZtFsccfk8pyJmp9LqwJjIq+Qt5vk7B4fyonhN5SBQXl9D5+e7XuYt5T/n vV/kWvaKE+dVVrDbI+csSziKlc7YAo56BDe9qFnDL6MWZEYn/Co8R9Y3Afz3Cyf2Cl3xklO3zXMR 55YpsDY9xRrO0mXoMlbJV8mf1lBogls3PbyEdoIgwUd+koByfGjhHTNoVeeunWXQIGB9FkQKWKJD 66ZsztnLmS36zSmiwDMy89fIEaL1MUoH1CJhwqYOB9TrRfrfm9VPi74xq48nu4kdPvaU4pSlYg8J 1KYEKnFKl0NKKAro6hsbYkpBV1AphKIUdfrGTohEAFPRsIlkZ2XXZTtEUI1lWbaM7NNRKSKhPrWD UC0SvpCBMEypY3cRK8c8sbcYUMf74MPxyfhcvBaD1jqQhCt17KpxOU4h9AkwlLAkZhJLCIg9g9Az ALFnD4k/iHaCQYdesYHSLfZ6usorNqii13cxaTszgdIttsqk/ZwJ3n9j0g5hAyXoRGXSpqGgF9rV pJ2Igm9N2kZc8K1JO58LvjVpT+OBribtS2TBAi95H3ha/AH80XwOH4qZAcjl2cM/xlfxEYG6AApY TRao+IggqYOQqOIjAhUhUcVHBCpCooqPCLpSPyDzA6gIiSo+IlARElUoMqAiJKr4iEBFSFTxEYGK kKjiIwIVIVHFRwQqQqKKjwhUhERVRx2oCIkqPiJQERJVfESgIiSq+IhARUhU8RGBipCo4iMCFSFR xUcEnZLPSj4iUBESVXxEoCIkqviIoNN/g5KPCFSERBUfEXT6b1DyEYGKkKjiIwIVIVHFRwQqQqKK jwhUhEQVHxGoCIl/SVk4DAmJosiVIiH5Jan3EwlNMza59BEJ8yNEWw4fiTgSIYpGMqKoIiHtd5FQ TSTcItpcE0EFYgUc4m6UKILj8DGyR3y3pDtRvTI2hWYKH7mxRNFq7SY7kVCvGuHY3VszorTE79FB YnXTFDk6RcxNrolYKubHEc0dzDsAvS4/SnaJfx6RK1ZLKxUDM/ErsUt8fQRVMkA8TJIg/n1z2gyR kJkRlZ6OgPPO02qjKMBcJCSsL0mImojbObdEldHLmzb3NMrY5Jwf0V207UOkKDYxM+J+xF7R5h73 hcsODfPnuB8IQJblg3sByZ/cD14JSHaAGi14m8FH9G7EZuqosIwISmh/OGG+IY2gdRgZzh3I0z58 iirUWZjW7j04I0I970z6GaFuSZ5j77ADh/LST1CP5am/OpmOnDqRd+JTfmgGLfxIjdPTXWopu4cX M55EUDLTq+5HILRpl6gzy9T2PaWmDHKQuhyYf6GGwklGQL+MTSHr0y+mswThyqHb07iX+y4hIbvj D1/qvbKY3WtlcVV+KeWArKDkC7XwSuSo+5tG3e/28GkcI496MGPuJa179lcM1NVu218hh7Z3kZll TvccBj1xvbIEG1oePT/F8GPkV25YfWJeldA9i/pMSAUI4nBfmE6+ne6nT27qLx11fuT96D9ELgjY O3yx5HYUQgmVOO4b2cg/Y4tjCenpmsL05xE5x6iFksLn9yOuv+K8lBC7wNnMCGG6fr6maJtDZoTr rspom7MfR71ibCwLvo/uKeI0o5z2wqehdQG7XgXsuo86iKJ+Ed6NKKJCVbP97ypuCiYo1eDrZcks hk5dU3F+acbVCqlNvcFbRkO+jBVT16yTZ1UmTJ97bbi6WmkUNWXgx6jWx7W3FXaEkrzrd2FvtxXD RKBJRBWPxaG/5nn4E7qXnNVMvyfuxTwsnjionilZxgRy8SbmG439TC2mtaSS+ZquCR3Uuw7uYdLH AdHPFUUTNnz/APdVa1mEs7d/wEr3EG90tBc/BPjl0o4iVgN0Tt/ueaDbAzVRRKJo8/r0StPzjGHA 1AWM6VmaexvJeGGQJTkj+V0CHkveSD7i2sRAgk3MvhhbusCNwVn6Vh5wsWbTGeA6Ybf0qLRQKpPW SlukajLQW5ZOTLrR02zutVN2nrIxrMHjxrba6FeYiSJpv1vslQmgZpE567wGXcISv7Uf61B3lmAJ 3FYsqrm3p+wOAcYsVLtzyi5sQcF048rVvC3LEOSU3cUFvy0tnX6VN7T8OgZUThI9cAt8JnsxDnx8 H1mG8xPrxlkryfFCJTleKZQc+Hsvrd8tP7R44F8nVI2437Y+kVRKS+Dca6ExMq6vjz7td8sWz95f uDZLb5bBWorSx/mRM89nWWjM+OSQWONKmrl4xn4kNGZDsv2+qNhYMfiW4HKWr9DJGCfZ8Tq9Qlll CbtyqLnnDcDvlj7HoS+ZP1Fp41xJr5h7zXhc2Q7R0CzWjIuDL4juXRG8dJkhXNvzKSb63CS6X3bA zTHXeCHyXE/ss2xc2dsrY68YV76TUTSqkXFlk+9pPrG5Yod9Z+ZeozRzt2A7XuOXoWbbVVwmEW6q UsFnpxC6Rf1BdvQRoRLwBkX6zhAfficeQ6FTtY05fm0R2H4ZK5P76aAG99MfBupbK0Z3q35vQwcu 9K6KLHfo4Dlmw3Zh0/AddJQ5ESd7FKHMevofzO5kiyJj1pI9CmERvTfZXz9AP87sxwwcOag9YsR0 71XeQe4hBD+IZRUQsNJP6uke4sdf5fp0CEiUeOOwFD6W5OBSvAJ3foa/ew3U39W02BkGlfYsb1XT CyotGvd+McWaOAyF/8qjkeRLL2LJQkXWNGRFk5VOEKy3jHWaU6Kqo84TGkb1OXZLDc7ZvzJXa4r+ TTMhPRLAukSssEouotb3ogxaQfYKERYPGKutQBDj8RqjizndViAxPcPVTmGMqw8eI9vpjf3Ks5ja 7NUDyCE/c/Jprg+zz4ItyFhmI5MBvcKvcvY2GKok4QAFC8ck9t6twUnY4UeeyGcT6fr15kidq54D ocPh5A6tmbziDqV60Zl5nrdWnukzClUobCl0L0HXSYclqIEsgiJyXLtlXNgcmWCtn9vJ2KEc3aDS 6W4bZOXRHWid+2UoS+A8uYdL1dtXa1CQtOGh7I+ycWFvV49drVv5TkDR2IGMC5scq7ndZrWdKDQO UK1e2dK5BrN+BWM4Ok0CDMRjuvxg7hWsF96IufPoOKNf/LxVbPGWQ8OgengX/+Z3RQpvLGc0+1qM NHeaX6RqrYWd8gGw5FOtKJvjLa+NOjVHYHHczcHt8+ILkQKgwOS1Vpg/2y9guWuf0nDTPoWZBXyJ u/BQ8EcZZ9Sp16dHntatfCMFt5BRp4TjryA3J59GZRAEYzSdSjka6QCRMJoSZY8dDBL7i63FjIdO h42xGXW2VrQVSFwqVdmDm2yXP4rCQwZcG+Zlv1gtyZaDtLIX9bw7gPGWpbWiQN6q1hNvVatpVcNQ 9IuELl0YiuYEbhvT88R1KetYXqubdtkxcgSWtLE1LTHzHAFY6CtmwtRXCU/vvre1evqR0u29GqTk QUZeg2wTnQrcGBicBGEO1GueZr3o1yYtrjUVefIeecEFSe9sXh4i+vEMmV8Iq4YBdDCavobpRN// 3p9+lrjGnNRiMiab7vyOvrg5aFHTb2kF7b11kYF6CBg23KR9/AjntQHewd4hRIjfKl+Ws/caIoTl t4o1je9JQFsxa970ObGzgZ//DIiNmfQ28cOkD1Elk9oUTHTpgCuQiT7IGkGEMZVWejvESldCVeLy Qjac/hmcY+B8ljFsZkh+0cSXTbpqh4fWnhiW5oYrMQDFp06ewFn3yDZ6uOnVGzhL1Zq9xFkai3PV lpkm5elFWum8Hrtn1Xgtf3Mt2uLIVeZqS9aYx5trqcUkzOj7qQRsYb4ZY2lnP2x8M+xPriK0hE2i hBm5X7TOGWez7blFl+jPjSdoDQ/dlHYd8UAo99SoiXMq2Wl2vU5KrBt0AL3iJRM2bgPYo9kvxR/E HypkAySjjxSfMmKaEsIhcM7l0Pa25M8TazbAGZdadZeXh8CZeU6HHZacZircV6+WbJHslQg7/aHX sG9zqQiFXmG5drePnRTpdcxhUguhzeEoJkjIsROHs/rR3P6t7I9aG6Ssi/Jlw9awLzzP21M4tldQ 0qmNRWt4xzLutTzJPbXxzhltK4bnotFe5sIetwdG7vZJ8dWNWiRYuREBu33SV/tvOOx7VBAqgyB0 cFRWKAvh1cpaZGoYjemDsSZh66POLEsUbUweoRBrV06WQvdKn9FgNvs4JWmI0D5LR3Pua/80s931 kIDRnd9V9hUMuvA6wuzsjqwTfGEM5nYw9aNxRnIDX6FnVzJ+tfjMau0TlLCzZx4ImbDhbbQQaJ2p P8nZkTX39PZD2iVcqftFZEcWSDrtUbjvUJq0NqCfaKSKb6OiXC6iO8goIQ9HvhGplGG5PJY776M0 mrdNvD+lJjupfOEnq5KishSFPiDV4OMzHhSK7SGBOrGmKqFYKuVY8S8f+zi+PCRJ2e+mkzYk+uUT yJRXdhK/vPoggXWBm3btqdYsNcv8P8adIXsIgCK3Pe34UqfVoW51iC5h1I4gji9dW4d8cavzJKB+ GoACavXSkVh3mVIUBMTLfGTjtMoXzqY/i82okBZT8m9QS2QKXPilzT1n7C6Hc6MNH2Y++Dium9vg oWlGnA8VLieT8AqXi/OLsRI2tLKRYxyzly482PCewHFWeS2sCsxQVpbuPHzLkZLn53HWq4ToqQNZ FXE+GRX7/EYNVdO6/U69QW6Ktb5oavjINWBwDVpuDkSkxWo96EOnDeuGuI2jvyvdSndDdnDdkFlP jx2k7DDQArOeIog5Ary6+VSEy35np6i/YLupmd8Fw9AV+hzUCU3W8EL691w3iXMhKR59wi6YssOg oWnMLXTmu1J6wM6gtuLj5e2TrAsRKxuoWeGtafLZNaHh3Wy+p/tKFqFF1g8eZPVAbsybN282K5gI WbvSO3i5tzc9JBj49aPBgX7Dh6bEhg+H3zYoagtBkLK2ON2sjQizaaOhwhukJhW9k0FHSc3EFIz7 OMuX2PveTmjA4cwnqnqcveRA5XAEL6Wjp/WOachP+fT7vIdb7uu3hWxMpw74eLoJUiOV7mXBR0wb H4gfYFq9Vnc0MznqI/8VH2Jb6Fg2njNHZt72enuh47FjA4Oi29LV03HW24Hc5y3rrzgE/jIKtuWt 5KjhApP19umRujkDr+qHjTtjWhujsaSq91Z3+5NHfVaszvbVXJm/UrBxI3LU5+RqwYZzvqBQsJuA Mu4yopaI4anRAeRNGnN/o/PoO6J+X5YlouzVYEMWNfzq3tABkwHntc2ZO35jzWcuZ65jgm3MKH4u 8xKzkklWQX1MDiBPewoGjCYizExis+DoGHQOjy1QHF2Mgk9GRITahMWn+qECdIj0AAo80HBpDAoJ hRdQurROKhwr1Zb14vaSTuTO5k6LsnzrMyucYy3gwEqcmM3y8vbxW+VHwB4Qy32V16CHxe6hK0M2 cyGK4atucM6B7rzeYhOxcJKY7J7yoOPZVN7xBSWY/sn8ibskcQ9N6RT9ck+Vtj+g8w34Y/m2/AV8 Pz7kVBWJ/XLSzUz035fwBUUDc/DeF1sSJXFRi1paF/2iV+/26/q5R79EDK011dC5uTN//Sc+wqIN T79T6Am/1z3L9pdGPW4fTTbjh65FaYV8vLjhveOaoEXvTnogQzxW67+vbNVr3RJyl6gjP1b99wM0 6r8Ytg6lQ6lDoNQ6xDGl1OEmGeu5QuuwGxPfbvaEthoDW7BY3YsVu/rmYh/vLtqtVksboNOjUjOn u1rtUttzd3eJzT/TAgbXeNejXvJ9O6FTrljdutkvXs1+EXH/SVK7VYO4Ce8hWfVe5ZXrBMFqZOjk GRTPlLA+L+LdoUirnm8kWHmD3QoGRxxYriPxeztULapRs2mgNfaGcgyhUISzwY7adMlJSanknqSC 3iYZKgM7pTNlp6Sh0rzxo1kX0FapZg16VUqbW4/KjLkM2WwuoPZezq1EPWT7uadRKLglH6x78ppW ZY4a/BClK0d7Ba/UcnUOQbz4uusud8o/uanMLR9n1AwayAaQ7KUUj9/C3svOYbev/FD7e+nEM0G1 oIL9bPS64q1HvvhAE7IZCmzQTjeKbdxfeIDJe4HmCLpxV3GfCVCuhgic4o4QPVS4UQwWYYJabtoR 0SXBTdEE0aQl/YuuLyFCg4OWBLh7riDcfb2DlwT7+qutXOM0NogXxYP+JM/xOJARVUAHuO4fMZ7G Q72HaA/1fj6xXp5QObs+eC3wxEPxWHw3fhQvxKG2vjCEr/Z5bTyydtCpUYTaLs5Ls+NCkzGHvAmz DH/+0G3LGx/yWWP3nvjybBufVTcj+OAE6u0/svh7zshaBp2qPj3wtGbzW0zadBEZdGrolebCUaeB qUzlTecg/aP4Kh08pP9B1xTPYQ5mIg9tDvfFjKIjp7p2c94r3DkO2loA1M6+JL7BU0pna4/qXzR8 qDeNo2wsiyCs5zd+Kh9qpLP2SfE1g+smvdksZfUmKzwt+ePGoTsqkLWg3/teAmPBRMFswTJBgGCT ACQKsgQ0DEoNGGG/E7oisAgD9BkiI7pUNIP+TJQhAvmiOPpnokFURseIBvp8+plJS86mzd9t57fK x5Gv5rnr9nil15lS7B7m9gprw6DXGX1/xrAufmf24Uq/M+0v6iuU3SvQindr366DrO1rPpawJRYQ QMkwVBIMoRsCYMvUpl9D2fREOoF60w254DF9OhcwIcFQxG1DNZhppdxq9BXXhTstVc95rw2x0tt9 VWjA6ACNtWD66aIsHXp6qs5ufaF25mFmlvjGwLQFVcmexQM4L5KLbWdO+FB6urj3kQdUVjvLUeve gYApiwY84nA2XkpUn3TQYaTHkoJeNt4T3JadVOe4ZVfIFKLoQ8elMaiU5Vk6SDDdG+nhg1Aaiis1 Ll/xodAzkR5ZCGVzmo7GZdyHkm2Ge1LDKCkrzDQu9w6jaNDKRoRRNq+r0LiMhlHC09RL5pG7p4o1 LvuGUWLI3XDy3OnxGpdFYRQ6LewAebBPHu55OoySnPbB6zq5SzXAPbEwCt9R43JLGGWLPi1MI4ay lbyyfwylTwCZKDuG0pMWZhOz4PY0jcsLYigX0zUur4ihnCierXFZEEPpSwtLiKHU1JIHD8VQ5sXj nudjKG8dSzQuy2IoKJnLZzGUQbNpZa0xFCYtTDuJ0t2KvJCZRDGghY1NoqSmT8M9pydRQotxz8VJ lKPFNzQu85MoQ2hlkUmUoTdoYclJFGNa2dEkikkJrUyaRAEVuOedJMoI8lBdEmUYLawtiTKyhBZG z6LwyZwZZlFGz6aFWWRR2LSyWWSc2Q3ck5dFMaOVBWdRxgBaWWzWwBO4596svq3bb7teduhrvk3X 9fIBD+oRH7WTPi7ISR/qFR/kig/1BbnzwoeK+yC4D5Ua5oJQw6i9w5DeYdQR5M6IMCoahqBh1KXk ztIwqm8Y4htGDSd3wsOoojBEFEY9QO4cCKOeDkNOh1Hvkjt3w6hYGIKFUVvInZYwqkYMohFDHRLj ggyJobJjEHYM1SYmHbGJCfMYQb5pamgMkh0aQxXEIIIYagK5kxBDPRSDHIqhnid3zsdQZTGILIb6 jNx5FkNtjUFaY6jaSUi2dhKVmYQwk6hjyZ2xSdTpScj0JOpicmdxEpWfhPCTqJHkTmQSNTkJSU6i FpA7BUlUaRIiTaLeIXfuJFHrkpC6JGobudOWRKVnIfQsqmEWkm2YRbXIQiyyqLPInVlZVF4Wwsui BpM7wVnU2CwkNouamYUczMwyaUWGvze2QPqaU028h4/brvPHx96VFUu264DtOjvRgxYrzEYEUg3Q wKWBYEjEArPwQE/H2jw9s7uBATqA72hbbBNh4RgaQegkRLwoBucj3jqiZq0R2vHdrcbGT48Hi+P5 8ZHxyfFt8QXx0vg78aAunp5umG6RPiudbxacHptOy0x/XRtZsXBpIMthlNkuHcCgWowfRIvvyaD6 ugyivXBH5g5Xv9VzEC1Nx0Sj2HYQDWSbmWjok4dWmHlq7CBPBSMCTTRGBPbTRAM9NahDyIilZARK Riwgz1xKrgGZaY0hEf00PR1NNBaYkRG1eSYa4eQRPfIMT0cy4i55SW1eP02qgYmGHjyDT556lzzD tthEw7+YjBgSYaLBJ0/1L/bUsIURNmSEDXlTC0dPDQt4j1AyIpSMaCMvXQczlkBGJJARL8hLXsBL zpMR58mIt+QZ5GIJDCpMNMjFsjXCU6M1gozQjjfRIBfL7lYmGt3h044lI8jFcjq5JhdL8o2aaJCL JZ9c82FEJLlBLpbJ5DoZRhSQG23khpRcF8CIO+QGuWNZR67vwAjyqAa5Y0lPN9EgF0tgSG6Qi+X7 JhMNcrEEs8gIcrHkkWsePCOY3CAXy1hyHQsjMskNcrF8XWuiQS6WAGkyoSIAAErHAoNO5yYALisb 5eMa5SPr3iJt7whlfTuSQyis3Im9OLeffSArnoYm7Ub5hQJBDTPuIgDqjfKeVqMoN8aCJAv1T+9c xjTK9RvluthnpO7tFw7eVFDDTL0I+gPrRrkT6DmjUT657q3mYwJKzABTyren2zbKl5Dnf75oQeYL ng/zQx6e/ewD0ih/R6t725OKBGDMU4QmANcKALjTBq6ROV6FMQX4RcU1HIBWtjQNAdUmYBC8w5dp jXIjPsb8RaoBABOluCBqRo3yPo8JmpYa+aTGj6GikaaLWaOcaJZTyYj6lp4DGuXdHxPgfdP7ppfv XMiUKY1y7mDNaUMp7i5TkPqW5iZHOkHMVIvz6VbTZoI0NC0frIkglN7AuU+jnLaWHqve0NTzwrMP I3pRXMLojfRhgKehTmY4WnMQeQvHUchprb5qoxAwYzgNTKAbA2e6q977Jt+BR+jb6IehluOr5r3P Ptj2ojiCN3QFUmck/NWmh/p5eQO45bzcmxxTKTYdyG7TKDPwTq9RTmZ0undwCBEa5M2y4wf5a7iH AJf15MM2yheSj/dJfvvlOxqV/PnGNsord5E/3/giALQa5S62OpTnvQHaH2gz1XlRKTXMmaO0EieB Iy79QXPT9FfN8A2Qr4D6WN5T91Vzz2cfzB/LQZfwj5eaqFNksmVFX0tNyFhwA5YaoHgW6nd3ozwm 6p+0tze1vyMvpRXUSJn6xYocW00dRfmcX8McZAHam9SsyBwyAG8AeQuyrHzGPkf0eEyoZh3r3nLa G+WPyVJ1uwhoK0rh/A72lFypaORLueGIMZks8tYWoRQ1WEIfK2WutL7Z+csEbMkEjE7VMPcXkwn8 8mcJDHDCmMjRQHU1FlmcyfuRhcGoGu8THkO+VfKm61rkc2vaEEAWPy01GvlmujXKfVvkgx8T4+h2 9D4L6R/bn7f20YyibqfXtxDGGalWjnHFT5kc49KTGdmUfXSJep1l8qW15Is5MxO7y3z5rmfd215k Stx7GHNio9wZfge7ejfK1xNgR2+alt6xBgIk9+zVRw9xwJihaY3ywEb5bTC0Uc5bOQ9j6rTIDRrl C6xBv0Z5yXTGmBZ541a1qTr9G+VrhvaVgkIy8/ta5NZtgPxM4kb3BYsaYsEMru7ARrmXM3aUmREL tDV0yW8skcvDmDmxoI/esEb5zQvpjJFSWs+5OiWJh2qYwtOMPnomjfK3t0sZW04zkvTI72/crgU1 zMTTjPQpukSjfPWAfTWv1N3AGKbOxVcMf7LIcBvl84+CyeSPSYALOiytT1xdVKJDftoLiu+RH4eG LvmlV+4h73FmPMtGQ3cKU0eUWsNsG8/qozfT4G01AWiXAOAzAAcB/9GAkNUthVyo5EIjFzVyaScI QrX8DP9vh3aC/N27tLlfywW51ESLmz9ylzOOxmuCEcNOPjAl4xLJRaPjOAfAawGYQS7kdwgWkosa uXgBZe0Z0HHueqBs1Td3rJ+TFw0i12SzCbzNxpqPNTdT7Ks+hX92W7WeTi7jbAGAz9WjI35bl2cS dCyqEEFXLizbOXbw7JG2ynzD68hosFJ1Ik35QPBb/aapgeEpDQiPUMF8YA/mkQuXjJoGgoA7WA38 wCrg23HabJiXaX+dF2HHwgI3usaif7ZNPp5iDd+zMZm3bTB/HHK7y30QxdtVxv95EHacLOy4Rgg3 QVFRESh6XATkv2cBeRsBDqalkfHwDELxn0xcuYL/iM68/Aw/w8/wPyPo/Qw/w8/wM/wMP8PP8DP8 DD/Dz/Az/Aw/w8/wM/wMP8PP8DP8zw0KM+f/alvn/yZT77/5WYHe/2XG9H9/Tro8/3/kVYDODfBD XGfo9ucXf5/jfz7znY/9dfO79LvB8F9mvnP1z+UAdDzFXz//P54++D76H0tftaHMyo/Xa/2D6ZN/ elCp1O/f33+RPvj+/X+Xfndt8hVo/+XFel9/MXL1C5mBf/IHAKpC8xfvX7ubXnf4G/zVxV2fQg/8 0kOZgX+8GH5X/n94/m7a2mT63f/i4s5Xp7wDlapIX+8fLwP/R9//D+WPTlft/bN1yN9d85f3+j59 0Hmbf0/6f1b/gL+/5l9I/18J/5PT/6YW/rFM/nUZ/eZj+POj4GtZ+fbG/z3P/+39/uaw3nfJ/1N1 xjf3+/4hurYF3x3v0jz90HL99fP/TZXw5+0P+PbPD3vg76/sCKoq9y/eZ7c/vX9n+/Nj+t8f7tLW /F3PA/xNo/Z9+l3aH9V9O1P5elgP/EnO/yr9v23Uvqu6v2l/vvnVlOW+47AqT1+v7JIs3O1Gtjnd Ol+/6qaqU8jDtI6jf9P+qB4HfH0bPzRPXd4/0Os6ytNS19LT1v6aw46rVA/UXUtLXVv7T99/l/an a/6+Hv+T5unP3n+3blp6qhfw4027dVNX7zj6Yyn+9jP6IZXvSwbovPKb82H5Vu98h98d01Pvpt2t y/N1fv9/WlN2fv/fb3a9UvXnh9Cl1ugMGn925v9R+Ovq71+pGf8T4Yev7h8+/K8d+tNE/uXjf5PC P/H2wf8v4R9P/7/M33/frf6l5P+jJflvPvJ/S/LfrP4D6Ssb8v9UBkCX5v8/kr6q+Vc1vv/Nn9Z/ nb6yIf/aWQD/5rIIlA35f6wA/vXI5N+Vgf/s96/3n67/foaf4Wf4GX6Gn+Fn+Bl+hp/hZ/gZfoaf 4Wf4GX6Gn+F/VIDaKPodhlxGl+2f4X9fUAPTAB94glDgD7zBKhAC5im2/YE7CAJrgT0Z50OeEaSI CQF+5PaqLtdbAgroqif0j6UJwOH/Rt2rfzb9/+7wPzl9qKMEtZugcl/wV/UjMp6h3NIEX1X9wIWO c2GwBN/oKwmhLhTUdkoFSg2lw0Apm5TfsQ+XPuTCm+9gP41nNdvZ1mmulbPtXKs5tvNAx7mMr8cd nGzn29sumAc4FFU8jJ5tP3eWtdU8W9CpIwXnBrqRa6uOe6ieSVW8ujP+2Tfyvyv8f9/urzcAAAAA AAAAAAAAAAAAAA== ------=_NextPart_01C52880.DC295B70 Content-Location: file:///C:/C9344E54/Paper14_files/image001.emz Content-Transfer-Encoding: base64 Content-Type: application/octet-stream H4sIAAAAAAACC+2dDZhT1ZnHT3I/kswkmTHz4fCURaYIgiIKYmUobhHUhalWiGB11BacsQiDMBhk sLEuahS3RmWXVPtso/YrUO2Tpe0aiy1pWdpGUJrFR3dWC43abc3jqsQVNDy1Nfue+5Hc3LlJbpI7 mcz0PZ7jPblf539/55z33HvPHV4TIcQPqVFKl0wgZImdZMML/YRM+wwhky+78nJCTOTbB8xkqpkQ K2xrgrRe2m+/C1IHIS9C3m4ieWH/BIbsP50l15Cl5GpIV8G6S8ltZDUZJGvJBrJG2m8qPZYI594m 5znIu2E5YKKlExLMZDLKc+/IlTX5EjjnWjjrelJv4YDmWkeJowa+fYzcIl/cPx8T0vsnXGTlGacG O6aeGsxQFiddZE4b2SqXsQrSTNiPbvPB8ZZn566iS+EckL4PeXreVyJzV9H9BqRtNPRtEao1Ww8m Rd4J9TBdsb4BUrNCq036PQjt50XYaRrkJ0r70sAr9lWe36WRh+ZFzoLUIeVpkKt9qmKfayCdL7Xb G1lCroWVP+SFNgNtcze/wkzTaraJ0LSbV7av6RrXyCnW0/NPkPK0bLkNmgu0QZNhbRDq47iLPMAf IHuPi3W98Gyxrg/C70nP/2Www3KAHNx3jAQeOSoshbD7s6sYWDzVSrbSfelSfeZ5e48JTGmgx9Ik l5FtTyTXnhjFvvI5TbB/JHKUzIUlPR9JicdPVxxPJuXao1qnnD8/5RJ+T3BNWUUZH4bf66Jkq7oN Ut6l+n+t2dPrptchXD8NVpHVwM+PkQWwHxU9833x+CPy8ftE9qdgXyPYCzxLsFfrlPN62Sv7Btiq be1SvkNVPx1Snqj68wpItBFeT/sodC4L7PxVTty2cuX2hXf3bF9It32FboPlFarayEjhBsgvJhvJ rWQAancDuQN+L6HXB+lqaSmnOSo7otZzJ6QrId0LZbbBDh+o9NBtCU4c34rpuRT09JLbQdPNoGhz Vs8cDV0TVZqUTGU7yRSxk4wOO8kWsJOsyk6+aRHt5HkNsp08p2GFmaY/WZoITec0aNlJpoCdZAvY SbZO+upCn9gH6FIOTIkzy33VKDupp6+qdcp5vX2VVdhJpl7HKMlO7v3J0YJ2kl6zkXZS4DnCYxSj w06yOu3kI1bRTn5iHW4n/9daWzvZA8liE/VsseXrodtusumxk6vBOq4mN0HygK1corCJaj3zithJ pU3qFJdQhSLTV6ECrgLAF4Hx/BmkNkW9KG2T1rmJ4tzKcoDzNmV9tSvyy+m98qZcuRdCmXv53G85 rIbUC2kZpOlSHb87IHVCcph8CYSsM4vb5H1NkuWh+/3aKSY1VwwFAzJCRsgIGSEjZISMkBEGZISM kBEyQkbICBkhIwzICBmNVUYlZ2aQETLSQyejFX0+GsV8YWT1xhcZ1ZBR7gpEEMqYIaVjHoNydh55 ZMioJoz0XlhGC58mzZK7qaIWKQMhIqM6YVTuNVcWSeEOWx0pZFRLRgb2nRGGiIzqmVGVfa0I6CLd qgpYyKjOGKnH2mpsdskrz7+8YvsrAzIag4yKPItUNoqXuhssuadaGDIaZUbZ/qP37Vo1t5RGP7JW /CSMjIxgpISih1Hp28uKbgsMfwuIjGrLSEVH+zVXWca1sg44ejMFyGjEbLYU9b1OLaMD1iUpZGQ8 I61RXXV55byi175LqPltATIaNUYFXoiqnkgKd8NiHbAkKWQ0zhhpDuZEm5cKmeZbjnJvC+pwDhIZ IaNq77Pzr2G44VBtKgKz8NcjZb91QkZjhlEZ0/k6n+o0h7wx3deQUaWMSvY1nR8ZqXqW5iZkNI4Y lTsXWoRXyRfe+ktERmOTkeYgXfKS9HTDshjV2/MaMhqxvlZZN0RG45bRSJHSBFfmdyPIaMy/8zca mb4vicbE/BoyqmjivNiDa5FN1X1zhIzGEaP8h95MXtRcqdykhSZTlwEZjQajvOdWxRr1I+1o3A0i I2SEjP62GFVms5ERMio3jN5bDmRUx4zUpMp5Us2MhYCMashIs+sV6YyZsROQ0WgwGn8BGSEjZISM kBEyQkbICAMyQkbICBkhI2Q0jhl9Oue3MaP0/Sj7y+WK+MvlVP4sv0j9b0LmEYaQz7OEtMFBRyDN h/Wn8Q9xTfw6zsYv4TLcNO4DjuPe5N5if8u9wD7L7WGD3GPsNu4e9iZuE7uI62Unc9eyn7Bu9ih7 FftTdhn7DXYF62FvhDVr2TnsHWwb62f/zHyPfZPZxx5mXmWfYz5in2bauCeZi7jHmOu4ncydkJ6E /AFY90fuaYbjn2Om8oeZxfybzJf5PzNb+Tb2YX4O+x3ezf6Y97C/4L/Bxvifsi/yR9nD/CfsIX4y d4BfxO3lb+Ke5rdx3+SD3L38s9x6/rfccv5Nbi7/AdfGZ7gP4Zpehmv7EUevUcunL6fy6Vsp2zOp 70ozIYOw8UvAeBsn+thcY/ZyPXC9esrW8mnMqd7tTdU4lp6T+jeeSf1yQtnXw8adsHyck3ziMtcD 769xy5nHOZpXHqflz5hT+Qwt5nc0AeXstxDSDaBmNtTO76hc7hIo9+yG3G+9fkf7QUg/+h1Fm46M kBEyQkbICAMyQkbICBkhI2SEjDAgI2SEjJARMkJG6FMTGRlEB31qIqPqOhf61ERGRncu9KmJjIxj hD41kZFuRuhTExnVQV9Dv6PodzR/rEWfmshohJ9F0O8oMhILQ5+ayKgyRuhTExkZ19fQpyYyGgGb jT41kVFVjNCnJjIylBH61ERGxjFCf5HICBkZfp+NPjWRUXXP/ehTExmNFCP0qYmMKmKEPjWRkXGM 0KcmMqptX0O/o8ioJqTQ7yj6HUWfmsioGkb6XrmhT01kVDkp9KmJjIxkhL4QkREyQkZj22YjI2SE PjWREfrUREZ1xUiz6xXpjJmxE5DRaDAafwEZISNkhIyQETJCRsgIAzJCRsgIGSEjZDSOGY2E39Hv MoTssxByFhx0SYPod3QKn7F28sesE/nnrC7+MSvHD1pPctda3+A+az3ETbbu4WzWndxJy2buD5YV 3CuWC7gXLM3cryzvsfstL7D/YdnNxiz3sXFLH3vUchn7jmU6+4nFwbZaTzLnWl9nlloPM2usUWa7 9SfMD6xPMy9YdzHvQWq2Pc1cYPsJs9IWZQZth5lHba8zz9hOMkdsDvZt23SWNFzGtjT0sVMa7mPP a9jNzmt4gV3Q8B57cUMzN7/hAm5uwwpuRsNm7lMNO7mGhj3ch7ZD3O9tb3C/sp3kdtk4frvNxX/F NpG/1NbJf9pGr1GP789iPjafbyRCnVzgJOQhc+18bMrlzoZyHzTnfpfjY3Md+thE+4WMkBEyQkbI CAMyQkbICBkhI2SEjDAgI2SEjJARMkJG6D8SGRlEB/1HIqPqOhf6j0RGRncu9B+JjIxjhP4jkZFu Rug/EhnVQV9DH5voYzN/rEX/kchohJ9F0McmMhILQ/+RyKgyRug/EhkZ19fQfyQyGgGbjf4jkVFV jNB/JDIylBH6j0RGxjFC34jICBkZfp+N/iORUXXP/eg/EhmNFCP0H4mMKmKE/iORkXGM0H8kMqpt X0Mfm8ioJqTQxyb62ET/kcioGkb6Xrmh/0hkVDkp9B+JjIxkhH7/kBEyQkZj22YjI2SE/iOREfqP REZ1xUiz6xXpjJmxE5DRaDAafwEZISNkhIyQETJCRsgIAzJCRsgIGSEjZDSOGY2Ej81fNxICkSx1 ErLDLPrYvMz5BdNi5zmmi51W01znW2SaM0banbuJ2fl18o5jI3nJcS15xrGYPOqYTbY4ppAVjglk jsNFmhxN5F17MzlkbyO77X9H7refTdbY55Ol9ivJLHsvabffCQK+SY43PktebxwiLzemyeHGCaaD jQsgfdl0uPEe08uN3ze93hgzHW/8k4nYGXO7/QzzLPs881L7VeY19pvM99tvM++232U+ZN9uftf+ oLnJ4TfPcfyTeYXjXvMWx1bzo4515mcc15lfclxqfscx02x2usztzg9N05xDcC0RuKadcG2bTPQa y/Wx6YblgEncJ5jJf9m2w5TNTr6E3EbWktVkPak0HCPkuIs8wB8ge2G58oxTgwvPPjVI6/0g/J70 /F8GOywHyI7fHCOBR44KSzkwJc680HdM8qZJhGNpksvomCqWQVvKnDbqmvNA9nx0P7rtqVay1QT7 //WDo2QuLOn5SEo8frrieDJJPF5go9Ip589PuYTfE1xTVlHGh+H3uijZ2rdFcGWa13aV/khlv6Md qn2UvkmVbX0F1QDpvxnRiemLsPNGi7ht5crtC+/u2b6QbrveIvaFKwq8P+0TznUb1OwG4oHaXU02 k43w2wPrL4ZEfbDOk9IcyT/qPKlvzpO2z1Fp0/KzquXP1QYbrmIJuR366c/42vlzlcv1QLl7+dxv 9OeKYyUyQkbICBkhAmSEjJARMkJGyAgZYUBGyAgZISNkhIyQUbnnRV+lyMgIOuirFBlV17nQVyky Mrpzoa9SZGQcI/RViox0M0JfpcioDvoa+nNFf675Yy36KkVGI/wsgv5ckZFYGPoqRUaVMUJfpcjI uL6GvkqR0QjYbPRVioyqYoS+SpGRoYzQVykyMo4R+uFERsjI8Pts9FWKjKp77kdfpchopBihr1Jk VBEj9FWKjIxjhL5KkVFt+xr6c0VGNSGF/lzRnyv6KkVG1TDS98oNfZUio8pJoa9SZGQkI/QxiYyQ ETIa2zYbGSEj9FWKjNBXKTKqK0aaXa9IZ8yMnYCMRoPR+AvICBkhI2SEjJARMkJGGJARMkJGyAgZ IaNxzKiUP1e+iD9Xvog/18+zoj/XI7zsz/UhbrFzHXexcwk31zmNm+bkuHbnW6zZ+QL7jmMP+5Lj MfYZxz3so45N7BZHL7vCcS07x+FmmxxXse/al7GH7CvY3fYb2fvta9k19jvYpXY/O8v+Pbbdvo8l 9lfZ440fsa83tnEvN17EHW68jjvYeCekJyF/ANb9kXu9keOPN07liX0x327/Mj/LvpVfan+YX2P/ Dn+//cf8bvsv+EP2GP+u/UW+yXGYn+M4xK9wHOC3OPbyjzqe5p9xfJN/yXEv/45jPW92LufbnXP5 ac42fq7zQ7iml+HafsTRa9Ty58qr/LlWyvZM6jsTNgzCSRxQwD9yoo/N03gvZ7N6dZWt9M15luTX k1e925uqcSw95zWwnEn1wI8eONliWAZhyQrrerjFoMFkDXI0rzxOWb5Sl9JnaDVMzoANH0NBn7IQ wtjEa5vCZ6ynWzLWka6PiRaxPs5UlD0Zyp1s8Y54W7gdCmOhzAGGEKdU9h1mq20902CrVVvog7Lf hhW7YHmOTWwLfczb1l2goY85x0bztWgLWtdiyb2rzjuWrqf6z4f0ORB1I4h+soGQH/KiD9/F1t38 txpoWs0uttK0m9fDc7ri/MrrkX020/XhzOj7bJ63V/SFTJe19tn8+08d0+2zWa1Tzuv12Ux5y+z5 OmFPr5teh3D9NFhFVtP/7yhZAPvRBjzzffH4I9Lx9Jop+1OwrxHsBZ4l2Kt1ynm97HlVPSh9P09X 9FUtP9q8Dj/aD0Fpx6HfLoN+283l+9Gm27q40n60l0F9U0/aa4T/D5BbiEewaRdBmiXZoFkaeZrO hnS6yu6odVI/1EnQ4gYtH4Ft/DeVTrrtcU7kU0jnDcK5VpOboGXeLOm7QKHDVYaeOyG9BmU+S8ds 4Obi8/XQbSZe9PldTM8y6Cm9ZDO5HbhRTRdJZckalLzOVmlStgPZzpuL2HlzgbHvMdC/Ba7jX6CS t0n3Qd9q+Cr3MNz3adlqs46xz1xg7DMXGPvug7Kvh5MdbBTrkY599zdeL9x73tf4OEfzpcY+s2qs UO4v2y66T7AObFfHTtEm0GWtx42/fnBU97ih1inn9dous+RHXm5zyjarZa/MJewV7Xc3SIZwD5xk oyW/39Ft11tK97ulZAP0tpuhtjcL/6d9Zx6kOdKyU+pPcr5Qv1O2sU5xCfjEayBNhCyAhjwfDr4a NLVJ+zKqdqp1bqI4t7IcOM02JZ92RX45pAc6cuXOgxMvt+R+y2GW1OeWSXVDbWuz1LAmk8OkCwS2 M+I2eV+T1PLofr92iqn+vl9ALagFtaAW1CKEaCjg9/R4erv73V293bOriUZoyaQSiXgsEYskouFE NJSMBJORgBz9EFNhnxRDXjGmQx4ag/3pYK8YDdOSVMREXkxm41BeTMSV0TAtwXwtBXQVVVQLLcNE DVckaek3Sks6pY6p1HBFQ2E/jSFfNmblGKUlrKUlT1cRRhIdaMeGaknGwslYKBmlMRENJiJiDIhx KByQuEh0/AIXSZGRWtIFYnFGuRYEvdwILTHQkqElC1zCBegENelk5RirZXgsjimfjtFaUvFIMhYp i05WDpjkirSEe2eHeyDOgBhyd0IMdndADEBc1OxfZPV1CXE28QpRi5SaTqVaov2LUuGAGJNhfzJE YyLkSwR9Q0EvxHjAI8ZYwANacnSiajpZSwRjVsVaIgF1Q0hlYzItxmQS5PhmZ8+gopNnFSvX0p2K BDXaZU5OSpRDtXRZ1W1HQSerCEb0irTEPIKWTDrs7vR3z4hFw6Ic+Onuao5AU5blgBZ/lzX/aG06 cJtRsZZoiGa8PQnPIpCTAgrQjXq7IA9y4G5G0JKCtgNNWTwqHvINj1k5lWqJe91pQUvc158MeCQ5 Q/GE1y3mg36PiEbQ0izxiIehjlIx+n/hHiuYpyVamZYhb086Sl//D/k9EW/vUNAnSoj1d4O6SG9X Fg1oCXTntCSg04GKaDBfS7oaLb6edIxqSQS8oCUZDYMEkONb1JkEIyboCvm9oAU6OtidrJZUTOIi iFJpCVWkJeHrTcciNBPygxbou5ABFf7uziTc9UbDIhpoQVSLuzOrJZ0cgqpUjFzQrXKxMi1+qgWg JKmtC4S9vQKs/kD3DNACa2JCTUHdgQ0MKbQIPSgDcmBoh0cHMcK9QvYZonwtSX9/Oh4FLdBeIMY9 7oTQfILuLlqkxx339kCE5jNMCx21gA5oSQ1FU/FoMi6MX9FwSSEFtKQCVEvhY3IGEEaIUM+MXB3J kQ6nwpCYTgylh2J6hBTU4imqJScHBqywrEU9tMtadAoppiUGcqQYiwgxDB0d7A6YQbDKQgwIWjSK ikVCopZ+Wq2Z6rTIsR/aTtLfC60ZOhf09SHaUtxgmCHCEAoRbjA0rKWgBRqu0c/TQ9FIuU/poIXa ZuOf7UGLIobjhWJEjCEx4vsX1DLiWkp9B2cqMkdpKjBH+TknIX0cIV1NhGyQ5igXO2/hLmxapzlH adIxR2kqMEdpKjBHeS6UvRROthGWD0pzlOc2LeU2goZzmx7kaL7UHKWpyBylev7otBZx3uZrbYT8 gK/N/FHfply5XihwF5/7LYfVkHpV80fvDuTmj/pBCE3TFfvK80d0v/qdP6rfb0sRATJCRsgIGSEj ZISMMCAjZISMkBEyQkbICBlhQEbIaIwyKvlXNcgIGemhU8RLpn6n7fXBFxnVkFExVyzlenMt0/Ur MhoXjMpwBlyWm7sy/U4VGTSQ0fhgNHIe8Ip4nxrWN5HRmGBkYN8ZYYjIqJ4ZVdnXynKyWKVTxjJJ IaNRuYesxmaX6aNVr5NFZDQGGZX0PV7uKF7qbrDknmphyGiUGWX7j963a9XcUhr9yFrxkzAyMoKR EooeRqVvLyu6LTD8LSAyqi0jFR3t11xlGdfKOmD9+x5HRpkKSel7nVpGB6xLUsjIeEZao7rq8sp5 Ra99l1Dz2wJkNGqMCrwQVT2RFO6GxTpgSVLIaJwx0hzMCzv3LUyqwtuCOpyDREbIqNr77PxrGG44 VJuKwCz89UjZb52Q0ZhhVMZ0vs6nOs0hb0z3NWRUKaOSfU3nR0aqnqW5CRmNI0blzoUW4VXyhbf+ EpHR2GSkOUiXvCQ93bAsRvX2vIaMRqyvVdYNkdG4ZTRSpDTBlfndCDIa8+/8jUam70uiMTG/howq mjgv9uBaZFN13xwho3HEKP+hN5MXNVcqN2mhydRlQEajwSjvuVWxRv1IOxp3g8gIGSGjvy1Gldls ZISMyg2j95YDGdUxIzWpcp5UM2MhIKMaMtLsekU6Y2bsBGQ0GozGX0BGyAgZISNkhIyQETLCgIyQ ETJCRsgIGY1jRkb7xmQhcxhWzmMJWd5GSBQOnA/rr2zzcp9vu5a7rO1CbkFbC3de2wl2UtsrbEPb T9kTrUH2tdb72GjrAPtE65fYu1qXsTe0Xs7Ob/0ce3rrAvZky8XsSy2L2D0t3ezDLV9kN7asYZe3 fJX9TMsOdlLLU6y15TfsR67/YZMuhjvqOpN7xXU5d8S1FpIf8ntg3Utc0nWC+8jl4q0t5/OTWq7g P9NyM7+8xctvbHmEf7jlSX5Pyw/5l1r+nT/Zspc/vfU5fn7rs/wNrT/i72oN8U+0PsZHW+/jX2vd xJ9ovY5vaPt7flLbp/nz2hh+Qdsf4Jr2w7U9wdFr1OP7U2sf6kPTDWnAJOaDmfwXbztyzkEnX0Ju I2vJarKeVBqOEXLcRR7gD5C9sFx5xqnBhWefGqRt4CD8nvT8XwY7LAfIjt8cI4FHjgpLOTAlzrzQ d0zyrEmEY2mSy+iYKpZBW82cNuqm80D2fHQ/uu2pVrLVBPv/9YOjZC4s6flISjx+uuJ4Mkk8XmCj 0innz0+5hN8TXFNWUcaH4fe6KNnat0Vwa5qtE0byG0rzrGI99VMq+yPtUO2v9Fmq7AMr6Ht3SBe2 EPIetP23Wgm5wSJuW7ly+8K7e7YvpNuWwDobrLuiwHvVVZC/nGyEmr4Vanoz/Hcz5Am5GNI8SNQ3 69VSfqKU5PUujXbnUPRptkifZgv4u33yNEJuhcb7r7BxUPJ3+73TNnGPujyabZ7V4e+WLeDvli3g 7/YhOIkbTnYElgHJ3+3DLjf0cw/3kCvA0Xwpf7dsGf5ube2in9nFcMDVltr4u32gI1fuJVDIckvu txxmSTyU/m6bO3L+brtAYDsjbpP3lf3d0v3q198takEtqAW1oBYhREMBv6fH09vd7+7q7Z5dTTRC SyaVSMRjiVgkEQ0noqFkJJiMBOToh5gK+6QY8ooxHfLQGOxPB3vFaJiWpCIm8mIyG4fyYiKujIZp CeZrKaCrqKJaaBkmargiSUu/UVrSKXVMpYYrGgr7aQz5sjErxygtYS0tebqKMJLoQDs2VEsyFk7G QskojYloMBERY0CMQ+GAxEWi4xe4SIqM1JIuEIszyrUg6OVGaImBlgwtWeASLkAnqEknK8dYLcNj cUz5dIzWkopHkrFIWXSycsAkV6Ql3Ds73ANxBsSQuxNisLsDYgDiomb/IquvS4iziVeIWqTUdCrV Eu1flAoHxJgM+5MhGhMhXyLoGwp6IcYDHjHGAh7QkqMTVdPJWiIYsyrWEgmoG0IqG5NpMSaTIMc3 O3sGFZ08q1i5lu5UJKjRLnNyUqIcqqXLqm47CjpZRTCiV6Ql5hG0ZNJhd6e/e0YsGhblwE93V3ME mrIsB7T4u6z5R2vTgduMirVEQzTj7Ul4FoGcFFCAbtTbBXmQA3czgpYUtB1oyuJR8ZBveMzKqVRL 3OtOC1rivv5kwCPJGYonvG4xH/R7RDSClmaJRzwMdZSK0f8L91jBPC3RyrQMeXvSUC+Q8Xsi3t6h oE+UEOvvBnWR3q4sGtAS6M5pSUCnAxXRYL6WdDVafD3pGNWSCHhBSzIaBgkgx7eoMwlGTNAV8ntB C3R0sDtZLamYxEUQpdISqkhLwtebjkVoJuQHLdB3IQMq/N2dSbjrjYZFNNCCqBZ3Z1ZLOjkEVakY uaBb5WJlWvxUC0BJUlsXCHt7BVj9ge4ZoAXWxISagroDGxhSaBF6UAbkwNAOjw5ihHuF7DNE+VqS /v50PApaoL1AjHvcCaH5BN1dtEiPO+7tgQjNZ5gWOmoBHdCSGoqm4tFkXBi/ouGSQgpoSQWolsLH 5AwgjBChnhm5OpIjHU6FITGdGEoPxfQIKajFU1RLTg4MWGFZi3pol7XoFFJMSwzkSDEWEWIYOjrY HTCDYJWFGBC0aBQVi4RELf20WjPVaZFjP7SdpL8XWjN0LujrQ7SluMEwQ4QhFCLcYGhYS0ELNFyj n6eHopFyn9JBC7XNxj/bgxZFDMcLxYgYQ2LE9y+oZcS1lJoTrmT+aGkbIX0cIZe2i/NIZmFO+BZu Yfvams0fXQRlL4GT3QHLr0vzR/Pal3B3tK/jLmr/Okfz1cwfyfO3bJ3M375xUJwXpct6nr9V65Tz eudvWcX8Ladj/pYtMX9L59Oeb4K2YSHkitMI2a2av6XbdlqEubyC87c3CPP5q8kgmUX6yEbSC7+X SHO0SxTztTTNUWgwq/RrtS+uTtrX5E6x3uiyntuXWqec19u+9LQpTsc3Ac+3i23qxg6NNtUutqnS 3wSsJRuEus21q4uleWS5PbVX2L7U898+OOgaWPFLsN23NtZu/lsuNwoF9jfmfuud/56P89+oBbWg FtSC8984/43z3zj/jfPfOP+N8984/43z3zj/jfPfOP+N8984/43z3zj/jfPf+P4FtZQ5/13u30TT +e90CyHPWOCcsOyxinMlH7dcY/24Za9Fz98Ma81/mwrMf5sKzH//Dsq+CzRMbCVkv0WcB/1dy37L xy13WSa27rfQfKn5b1OR+e/puTme7N8yWwrMVZpqPFfZsVOcA6TLWs9VRiL65yrVOuW83rlKi2Iu nFfUA18n9UAZ0GsSWNBgFbmd2nSMLID9aAOf+b54/BHpeHr9tB5Owb5G1IPAtkQ9qHXKeb31wOuY M7bo+A5hHvTZ1Y2E/BH67JHG/Dljum1fY+nvEBaTjeRWMgC1u4HcofgO4WrFHLHWPLFaTwDS27Dy NSgTWJMl9nw9dNtcu3hMIT19wrluFvSsF/6y/WZpJntz9m/br1bMY89TzGUr57QnFpnTrsZOPwKN 57uQUmD0nmbEc+9kdjHvcLsYPXa6mrJ/1SiWPcWZK/v5xl3MGc6RL/tFl1j2OW25sv/TtYs5q01f 2Vr7mBX2x1wn36wc3Cf2a7rUOw7I9seob1b02B+1Tjmv1/4o2ZvqhP2wMViy/VvvLWz76TUbYftl 9gLPER6DTYoxmNExDph1fDt0TxMh/wXLy04j5FJTvt2l2843lf526B/A1m4Q/h0Ranc3Cv+eyBdU Y0Cnytaq/z0RtbYdkDZC+VfCiutA2z6VNrrtB7CiscSYcAmMB3REWEt6BXVrQd8GYezK/66pXRqn ZI3K755K/bsnTBG7yGjYRVqHP4ebZDsj3rcPMOJ+v2TXM79km5mPW5qZQv+WT6l7d6bAvTtT4N79 CdDRBxvfgOUuRrx3f4PtAx27mCdgSfOl7t3V321VykV+npG5eBn5eYYyGawZkzekOmlvzTFpbx0A DbuYN1oGGJqvJZMUrHgALihtzzE5YfebT9hrx+RVu8jE6cgxcToGmBP2Xcyr9gGG5stlQj+Va5XK b5by/w/j43EmoMICAM== ------=_NextPart_01C52880.DC295B70 Content-Location: file:///C:/C9344E54/Paper14_files/image003.emz Content-Transfer-Encoding: base64 Content-Type: application/octet-stream H4sIAAAAAAACC+2dCXwb5Zn/X2neGUm25Di+cGwT4gIhCTSQAxJzhgQKMVCCSVpiAk1iQwIOxEEh TjBlQzHHFoewjYD0UwG7sEoofLT5lyKOFrU0sOJKTfqHukASLUd39SksiNu0ULTPO4c0Gs9II2ms yO4zed/Mq7ne33zfe97xPDZCSD/4cvBu8FObCdnDAvIyBDvb5hIy6czzvkOIjWz+wE7mcIS4YN84 8Gvk456pAl9PyMvsOjaStjwzgSPPHELJ98hCciH482HbGeQasoL0kCvI1WSVfNyRsgYnIZuVMA/h Nlh321jshPgTiYT62nem4pp0OlzzCrjqGlJqy27drZ4sZ91GD5DVCkPugOg/+rSKLDnsy576I7/s STAWn1WRmbVkoxLHcvB74Di2zw/nOx6fvZytRUDgn4Ywu+5rodnL2XG3yfvY0rmBoU+lg00VroB0 mKLaXga+UqXVJf8+D/LPd+HGJkO4ST6WLYLqWPX1q3TCdvBHga+Xw2xRkv1I1THfA3+cnHcvpYRc BBsfEcQ8A3lzp7DYzvxKOo4wv1NQ568pOvfIq7az60+QwyxuJQ/aDfKgzbI8COnxYRW5TdhNnvxQ Sut506S0fgF+T3z+6556x26yWjhAfFv3iWtx2XnSciiW5Oc1ZCM7lq21V37om/0iU7awc5lX4kjm J5LKT5zqWOWaNjg+FNpHZsOaXY/EpfOnqM4nE1P5UatTCR8XrxJ/T6g6fDljvAd+XxkmG7V5kPHO Vv6tY7/fFHt23+w+xPtni1Nm1XSAnAzHbT6WkGM+ks7fK5/P7pmx/xKOtYK9yDMLe61OJWyWvbps QJHeXCeH6zXpUy+HiaY8LwbPMuHFrIxC4XLAwdfx0r4lS26dd2P7rfPYvsvZPlifq0mNhLwsg/AC spZcRbohda8mm+D32eChWYL2RForfqamHtHquZ7VUeBvgjhr4YBPNHrYvigv1hEZ9ZwBejrItaDp MlC0Pqlnpo6uJo0mNVOlnuQy1JOciXqSGtSTVFNPvuOQ6sljy5R68uiyxXbm/8cxjjB/dJlePckZ 1JPUoJ6kJVJWt98klQG2VhYuy5WVsmpVPWmmrGp1KmGzZZWq6kmuVNsouZ588tF9hvUku2cr60mR 5wi3UZyJepKarCe3OqV68hvn8HryPWdx68l28A6XpGeDK10P27fSZaaeXAG14wqyErwX6sqzVXWi Vs/cDPWkuk5qltaQhBLT1yEBzgfAc6Dy/DX4WlW6qOsmvWsT1bXV8QDnzer0qlOFL2B95XWpeI+H OJ8SUr+VZQX4DvCLwE+R0/h/u+VCSPaQS0FIl13apxxrk2sedtxzFZLXcsXFcEFGyAgZISNkhIyQ ETLCBRkhI2SEjJARMkJGyAgXZISMRiujrDMzyAgZmaGT0HN9fcxJYWNkpcYXGRWRUeoOJBBqlyDZ XRqDXA4eeWTIqCiMzN5YQg+fLs2sh2mcHikLISKjEmGU6z3n54hxgS2MFDIqJiMLy84IQ0RGpcyo wLKWAXSGYlUALGRUYoy0bW0hdXbWO0+/vUzHqxdkNAoZZRiL5NeKZ+sNZj1SKwwZHWRGyfJj9ula IV1Kq4eseY+EkZEVjNRQzDDK3r3Mq1tg+VNAZFRcRho6+o+5cqpc8yuAB2+mABmNWJ0tO3OPU3Mo gCVJChlZz0ivVdfcXi6P6PV7CUXvFiCjg8bI4IGoZkRiXAwzFcCspJDRGGOk25gTfV4aZLpPOXLt FpTgHCQyQkaF9rPT72F4xaHZlQGm8dsjOT91QkajhlEO0/kmR3W6Td6oLmvIKF9GWcuayZeMNCVL dxcyGkOMcp0LzcAr6wNv8zEio9HJSLeRznpLZophToxKbbyGjEasrOVXDJHRmGU0UqR0weX43ggy GvXP/K1GZu5NolExv4aM8po4zzRwzbCrsHeOkNEYYpQ+6E2kOd2N6l16aBIluSCjg8Eobdyq2qId 0h6M3iAyQkbI6B+LUX51NjJCRrkuB+8pBzIqYUZaUrmMVBOjYUFGRWSkW/QyFMbE6FmQ0cFgNPYW ZISMkBEyQkbICBkhI1yQETJCRsgIGSGjMczoWym7jQm17UfFXi6fwV4ur7Fn+X1mfxMCWzlCzqGE 1MBJe8GfyGyWC3fw44Qu3iWczSf4yfwnvMC/zcfo7/mX6OP8Lurnt9PN/E10Jb+Ons530En8Uvp3 2kbfpOfTJ+giehddQr30EthyJZ1Jr6O1dAv9G/cgfZt7mu7hXqdPcV/Qh7k6/n5uDn8Pt5Tfxl0P /n4I74Ztf+Yf5gThKe5IYQ+3QHib+4HwN26jUEu3CjPpA0IbfVTw0t8Id9HnhSfoy8KbdI/wd/qi MIl/Vjidf1JYyT8sbOZ/Kvj5PuFxfo3we75NeJs/XviErxMS/OdwT6/Cvf2CZ/eoZ9OX19j0zZft Ecx2pZ2QHth5KTC+kZdsbK6y9/Lt3A9Nxa1n05jXPNs7Uudcdk1m3/gYZpcT4l4GO7fB+j5etonL LQPeP+Qv4O7jWVh9np49Y15jMzST3dEoxPOMg5BWAHVMWfHsjirxng3xTitL/TZrd7QL7Y5inY6M kBEyQkbICBdkhIyQETJCRsgIGeGCjJARMkJGyAgZyddFm5rIyAo6aFMTGRVWuNCmJjKyunChTU1k ZB0jtKmJjEwzQpuayKgEyhraHUW7o+ltLdrUREYjPBZBu6PISIoMbWoio/wYoU1NZGRdWUObmsho BOpstKmJjApihDY1kZGljNCmJjKyjhHai0RGyMjyfjba1ERGhY370aYmMhopRmhTExnlxQhtaiIj 6xihTU1kVNyyhnZHkVFRSKHdUbQ7ijY1kVEhjMw9ckObmsgof1JoUxMZWckIbSEiI2SEjEZ3nY2M kBHa1ERGaFMTGZUUI92il6EwJkbPgowOBqOxtyAjZISMkBEyQkbICBnhgoyQETJCRsgIGY1hRiNh d/QBjpCnHYQcBSfNL5Psjh4uJJzNwn5nk/CUs0rY7uSFHudn/EXOt/iTnC/yzc5dvMu5jf/MsZ5/ 17GY/6NjFv+So5J/1vEBfcbxEv2dYyeNOG6mrzg66T7HmfR9xxT6jcNDa52fcdOdUW6hcw+3yhnm bnU+yj3kfJh72bmD+wD8eNfD3GzXo9z3XGGux7WHu8cV5UKuz7i9Lg99zzWFkrIzaXVZJz287GZ6 bNlOOrfsJXpK2Qf01LJK/qSyWfzsssX8tLL1fGPZNr6sbBf/uetF/oDrLf5Z12f8Dhcv3OqqEla5 moQzXc3Ct1zsHs3Y/sxkY/P5ciKmyawKQu6wF8/GphLvDIi33576jTY2sf5CRsgIGSEjRICMkBEy QkbICBkhI1yQETJCRsgIGSEjZJTrddF+JDKygg7aj0RGhRUutB+JjKwuXGg/EhlZxwjtRyIj04zQ fiQyKoGyhjY20cZmeluL9iOR0QiPRdDGJjKSIkP7kcgoP0ZoPxIZWVfW0H4kMhqBOhvtRyKjghih /UhkZCkjtB+JjKxjhLYRkREysryfjfYjkVFh4360H4mMRooR2o9ERnkxQvuRyMg6Rmg/EhkVt6yh jU1kVBRSaGMTbWyi/UhkVAgjc4/c0H4kMsqfFNqPREZWMkK7f8gIGSGj0V1nIyNkhPYjkRHaj0RG JcVIt+hlKIyJ0bMgo4PBaOwtyAgZISNkhIyQETJCRrggI2SEjJARMkJGY5jRSNjYfK5c2rawgpB/ sUs2Ns+s+K5tQcXRtlMqnLZZFTEyuSJCait2ElvFj8l7nm6y13MRecyzgNztmUk2eI4giz0TyExP FRnnqSTvuyvJi+5astM9kdzinkYud59IFrrPI9PdnaTW/UOI7Kfkg/LHSbR8kLxaPkReLm+wvVB+ CvjltpfLb7K9Wh6wRcsjtg/K/8dG3NRe655kn+5usS90n2+/3L3Sfov7GvtO9w32F9232t9399vH ebbYZ3p+bF/sucm+wbPJfreny/6YZ6l9r+cM+3ueY+y2iip7bcXntskVg3Avj8M9+eDe1tnYPeZq Y7MN1t026Rh/Iv1h2522ZHDS6eQacgVZQdaQfJcDhHxYRW4TdpMnYb3ksC975k37soel+wvwe+Lz X/fUO3aTRcIB4tu6T1wrC5flyttv2i9b0yTiucwrcdQfKcXBcsrMWmaac3fyeuw4tu/nNWSjDY7/ +yf7yGxYs+uRuHT+FNX5ZKJ0Plu0OpXwcfEq8feEqsOXM8Z74PeVYbKxc4NoyjQt76rtkSp2R+s1 x6htk6rz+mKWPuD/xElGTF+Gg9c6pH1Lltw678b2W+exfRfDNigO5FyD56ed4rWugZS9mnghdVeQ 9WQt/PbC9lPAMxusc2U/U7aPOlcum3Pl/TM12vTsrOrZc3XBjvMpIeuhnP5aKJ49VyXeayDep4TU b7P2XK9Ee67YViIjZISMkBEywgUZISNkhIyQETJCRrggI2SEjJARMkJG8nXRVikysoIO2ipFRoUV LrRVioysLlxoqxQZWccIbZUiI9OM0FYpMiqBsob2XNGea3pbi7ZKkdEIj0XQnisykiJDW6XIKD9G aKsUGVlX1tBWKTIagTobbZUio4IYoa1SZGQpI7RVioysY4R2OJERMrK8n422SpFRYeN+tFWKjEaK EdoqRUZ5MUJbpcjIOkZoqxQZFbesoT1XZFQUUmjPFe25oq1SZFQII3OP3NBWKTLKnxTaKkVGVjJC G5PICBkho9FdZyMjZIS2SpER2ipFRiXFSLfoZSiMidGzIKODwWjsLcgIGSEjZISMkBEyQka4ICNk hIyQETJCRmOYUTZ7rkIGe65CBnuu51DJnuteQbHnege/oKKLP6XibH5WxWR+coXA11bEqK3iJfqe Zxfd69lOH/PcRO/2rKMbPB10sWcpnelpo+M859P33Yvoi+4ldKf7EnqL+0p6ufs6utC9hU53P0hr 3U9T4n6dflD+BY2W1/Gvls/hXy5fyr9Qfj34+yG8G7b9mY+WC8IH5UcKxL1AqHX/QJju3igsdG8V Lnc/INziflTY6f6N8KL7eeF998vCOM8eYabnRWGx51lhg+dJ4W7Pw8Jjnp8Kez19wnueNYKtok2o rThemFxRJ8yq+Bzu6VW4t1/w7B717LkKGnuu+bI9gv2GHT1wEQ9E8E+8ZGOzUujlXc5eU3GrbXMe Jdv1FDTP9o7UOZdd83uwPobpgR/tcLEFsPbDmorb2vkFoMHm9PMsrD5PHb9al9pmaCFMDoMdX0FE jQ5COJd0b4cLCechjoRzpNOjySGlxxGquCdBvJMcvSOeF66FyCjE2c0RUiHHvcnudK3hylzFygud EPdfYMMOWB/tkvJCJ/cX5w7Q0Mkd7WLhYuQFvXtxpJ5Vp53LtjP9x4E/DURdCqLvLyPkEUGy4bvA uVP4WRnzK+kCJ/M7BTM8p6iur74fxWYz2x5MjKTN5v2mbDY/9M1+0f4xWxfbZvOBxv2mbTZrdSph szabGW+FvVAi7Nl9s/sQ758tTonVlI/3kZPhuM3HQrn6SDp/r3w+u2fG/ks41gr2Is8s7LU6lbBZ 9oImHdS2n6eoyqqeHW3BhB3tLRDbh1BuF0G5beXT7WizfS18djvaiyC9mSXtVeL/3WQ18Yp12hzw 0+U6aLpOmPlp4A/R1DtancwOdQy0tIGWL6Bu/A+NTrbvXl7iY6RzmXitFWQl5MzLZH2zVDqqctBz Pfg3IM7HWZsN3KqEdD1sn02QbH5n0rMISkoHWU+uBW5M0xw5LkWDmtc0jSZ1PlDqeXuGet5u0Pbd A/o3wH38BBJ5s9wP+lnZdfwd0O/Tq6vtJto+u0HbZzdo+26GuC+Gi71QLqUja/tuKb9Y7HveXH4v z8LZ2j67pq1QH6/UXewY/4jWXQdM1V2b+QNiPcDWxW43/v7JPtPthlanEjZbd9llO/JKnlPnWb36 yp6lvroN/FC5VBFWw5horSO93LF9F8O2sgzlbjmEzyBrodzNh//Xki4oXd1kEzkbts8FP5P1W+Sw 8rtJFTYqg+r81iytAaV0Px8fSshUyNRHw455oK9WPpbT5Fm9axPVtdXxwGU2q1nVqcIXMFb1qXin QrynOlK/lWW6XP4WyenE6tlKOZNNIntICwis46R9yrE2ORey456rkHzpvcuAWlALakEtqEVcwgFf v7fd29Ha1dbS0TqjEGeFlkQ8Gh2IRCOhaDgYDQdiIX8s5FNcP7h4sE92gV7JDQW8zPm7hvwdkrNM S0zlomkulnSDaS46oHaWafGnazHQlVFRMbQMEzVckaylyyotQ3Gti8eHKxoM9jMX6Eu6pByrtAT1 tKTpysBIpgP52FItsUgwFgnEwsxFw/5oSHI+yQ0GfTIXmU6/yEVWZKWWIQOXmVEqB0Ept0JLBLQk WMwil6ABHb8unaQca7UMd5kxpdOxWkt8IBSLhHKik5QDVXJeWoIdM4Lt4KaCC7Q1g/O31oPzgZtf 2T/f2dciuhmkV3R6pLR08tUS7pofD/okFwv2xwLMRQN9UX/foL8X3IDPK7mIzwtaUnTCWjrJmgja rLy1hHzajBBPutiQ5GIxkNM3I3kFDZ20WjF/La3xkF8nX6bkxCU5TEuLU5t3VHSSiqBFz0tLxCtq SQwF25r7W6dGwkFJDvxsa6kMQVZW5ICW/hZn+tn6dKCbkbeWcIAFetuj3vkgJw4UoBh1tEAY5EBv RtQSh7wDWVk6ayDQN9wl5eSrZaC3bUjUMtDXFfN5ZTmDA9HeNins7/dKaEQtlTKPgSCkUTzC/hf7 WP40LeH8tAz2tg+F2VTAYL831Nsx6O+TJES6WkFdqKMliQa0+FpTWqJQ6EBF2J+uZagQLX3tQxGm JerrBS2xcBAkgJy++c0xqMREXYH+XtACBR3qnaSWeETmIorSaAnkpSXa1zEUCbFAoB+0QNmFAKjo b22OQa83HJTQQA5iWtqak1qGYoOQlKqWC4pVyuWnpZ9pASgxVtf5gr0dIqwuX+tU0AJbImJKQdpB HRhQaRFLUALkQNMOQwfJQV8hOYbIXUusv2toIAxaIL+AG/C2RcXs429rYVF62wZ628FB9hmmhbVa QAe0xAfD8YFwbEBsv8LBrEIMtMR9TIvxOakKEFqIQPvUVBopjjWnYpM4FB0cGoyYEWKoxZtRS0oO NFhBRYu2aVe0mBSSSUsE5MguEhJdEAo61DtQDUKtLDqfqEUnqkgoIGnpYsmaKEyL4rog78T6OyA3 Q+GCsj7IckobVMzgoAkFBx0MndpS1AIZ1+rx9GA4lOsoHbSwutn6sT1oUbnggJELSS4gOXz+glpG XEu2d+JsGeYrbQbzladWENLJE9IyjpCr5fnK+RWr+ePHXak7X2kzMV9pM5ivtBnMV34b4l4IF1sL 69vl+cpvj1vIrwUN3x53O8/C2eYrbRnmK7XzR5UNhKyESHoB2oBQnPmjznWpeDdBvHuE1G9lWQG+ QzN/9L/dqfmjK0FIl13apxyrzB+x40p3/qh03zNFBMgIGSEjZISMkBEywgUZISNkhIyQETJCRsgI F2SEjEYpo6x/YYOMkJEZOhksZpo34F4afJFRERllMsuSq2XXHM3AIqMxwSgHw8A5mbzL0QZVhkYD GY0NRiNnDS+DJaphZRMZjQpGFpadEYaIjEqZUYFlLSeDiwUaaMyRFDI6KH3IQursHO21mjW4iIxG IaOsdshzbcWz9QazHqkVhowOMqNk+TH7dK2QLqXVQ9a8R8LIyApGaihmGGXvXubVLbD8KSAyKi4j DR39x1w5Va75FcDSt0OOjBJ5kjL3ODWHAliSpJCR9Yz0WnXN7eXyiF6/l1D0bgEyOmiMDB6IakYk xsUwUwHMSgoZjTFGuo25saFfY1J5dgtKcA4SGSGjQvvZ6fcwvOLQ7MoA0/jtkZyfOiGjUcMoh+l8 k6M63SZvVJc1ZJQvo6xlzeRLRpqSpbsLGY0hRrnOhWbglfWBt/kYkdHoZKTbSGe9JTPFMCdGpTZe Q0YjVtbyK4bIaMwyGilSuuByfG8EGY36Z/5WIzP3JtGomF9DRnlNnGcauGbYVdg7R8hoDDFKH/Qm 0pzuRvUuPTSJklyQ0cFglDZuVW3RDmkPRm8QGSEjZPSPxSi/OhsZIaNcl4P3lAMZlTAjLalcRqqJ 0bAgoyIy0i16GQpjYvQsyOhgMBp7CzJCRsgIGSEjZISMkBEuyAgZISNkhIyQ0RhmZLVtTAqBPbBx LiXkglpCfgMnngjbz6vt5c+pvYg/s/Z4/qTaan567af00NrXqKv2CfpJzb30jZqb6dM13fS+mh/Q G2oW0WU136En1pxGD6k5mX5WfQr9Q/V8uqv6HHpH9ffp2upV9ILq6+gJ1XfSidUPU2f1f9Ivqt6l sSqO31d1BP9a1Vn83qorwG+B8C7Y9v/5WNWn/BdV1YKzeoYwsfpc4YTqy4ULqq8X1lZvFe6ovl/Y Vf2I8Ifqx4TPqp8SDqn5lXBizRPCspr/J9xQs0O4r+Ye4emaW4Q3atYJn9QsFVy1pwmH1n5LmF7L CSfVvgP39Du4t/t4do9mbH/qHcNsaLaB77ZJYX8i/cHbnSnjoJNOJ9eQK8gKsobkuxwg5MMqcpuw mzwJ6yWHfdkzb9qXPSwPvAC/Jz7/dU+9YzdZJBwgvq37xLWycFmuvP2m/bJlTSKey7wSR/2RUhws 18ysZWY6dyevx45j+35eQzba4Pi/f7KPzIY1ux6JS+dPUZ1PJkrns0WrUwkfF68Sf0+oOnw5Y7wH fl8ZJhs7N4hmTZNpwsl2Q1mYqtKBjng67DeVDowBuyeRBVucErfNc/aTk+G4zccScsxH0vl75fPZ /bN0+BKOtSIdRLZZ0kGrUwmbTQeqCjMbsYot2HpNWqntxarrn8Xgrwc/uZqQo6CgfV5DyN28tG/J klvn3dh+6zy270e8ZNf2XINn2ssgfD7pJpeRq0krpG8nhL4L2+aq/EzN7yqdsu4p0Mbw/eMJuQq0 /hR29sg2hh8cv46/q8pbNBvDW+AibXCxV2Dtk20Mb6lq418BDVuqfDwLW2lj2AmJfjJEMh9OuNBR HBvDt9Wn4p0HkVzgSP1WlukyD7WN4cr6lI3hFhBYx0n7lGMVG8PsuNK1MYxaUAtqQS2oRVzCAV+/ t93b0drV1tLROqMQZ4WWRDwaHYhEI6FoOBgNB2IhfyzkU1w/uHiwT3aBXskNBbzM+buG/B2Ss0xL TOWiaS6WdINpLjqgdpZp8adrMdCVUVExtAwTNVyRrKXLKi1Dca2Lx4crGgz2MxfoS7qkHKu0BPW0 pOnKwEimA/nYUi2xSDAWCcTCzEXD/mhIcj7JDQZ9MheZTr/IRVZkpZYhA5eZUSoHQSm3QksEtCRY zCKXoAEdvy6dpBxrtQx3mTGl07FaS3wgFIuEcqKTlANVcl5agh0zgu3gpoILtDWD87fWg/OBm1/Z P9/Z1yK6GaRXdHqktHTy1RLumh8P+iQXC/bHAsxFA31Rf9+gvxfcgM8ruYjPC1pSdMJaOsmaCNqs vLWEfNqMEE+62JDkYjGQ0zcjeQUNnbRaMX8trfGQXydfpuTEJTlMS4tTm3dUdJKKoEXPS0vEK2pJ DAXbmvtbp0bCQUkO/GxrqQxBVlbkgJb+Fmf62fp0oJuRt5ZwgAV626Pe+SAnDhSgGHW0QBjkQG9G 1BKHvANZWTprINA33CXl5KtloLdtSNQy0NcV83llOYMD0d42Kezv90poRC2VMo+BIKRRPML+F/tY /jQt4fy0DPa2D0G6QKDfG+rtGPT3SRIiXa2gLtTRkkQDWnytKS1RKHSgIuxP1zJUiJa+9qEI0xL1 9YKWWDgIEkBO3/zmGFRioq5Afy9ogYIO9U5SSzwicxFFabQE8tIS7esYioRYINAPWqDsQgBU9Lc2 x6DXGw5KaCAHMS1tzUktQ7FBSEpVywXFKuXy09LPtACUGKvrfMHeDhFWl691KmiBLRExpSDtoA4M qLSIJSgBcqBph6GD5KCvkBxD5K4l1t81NBAGLZBfwA1426Ji9vG3tbAovW0Dve3gIPsM08JaLaAD WuKD4fhAODYgtl/hYFYhBlriPqbF+JxUBQgtRKB9aiqNFMeaU7FJHIoODg1GzAgx1OLNqCUlBxqs oKJF27QrWkwKyaQlAnJkFwmJLggFHeodqAahVhadT9SiE1UkFJC0dLFkTRSmRXFdkHdi/R2Qm6Fw QVkfZDmlDSpmcNCEgoMOhk5tKWqBjGv1eHowHMp1lA5aWN1s/dgetKhccMDIhSQXkBw+f0EtI67F 6nl4Nn+0sJaQTp6QBXWEXC3PH51Xu5o/re7Kos0fzYG4F8LFNsH6dnn+aE7dQn4TaJhTdzvPwoXM HylztbYSmTPfzEtz0Wxtds5cmau1as7czFytVqcSNjtXa1PNmfMm5m1tWeZtfeDPGidN1u0ZT8ha R/q8Ldt3sUM6x2jethPCZ5C1pIPMh//Xki4ynXjJVWQVpPrZqnnbBZp53CZ5faG8rU6lz665N728 x5fqewIlmvcKfU/ATH7js+S328C/CCdOhjy1FA68S5Pf2L4bYVtZhvy2XHxX5DLIY9dC2q6H3HYN /H95Wl5T8lOT7JtVYTN5TDs/vgoqwq2w4d/gAoHy4s2PK/HeD/E+WJ76ncv8eC3Oj6MW1IJaUAvO j+P8OM6P4/w4zo/j/DjOj+P8OM6P4/w4zo/j/DjOj+P8OM6P4/w4zo/j8xfUksP8uDPD/LjTYH48 0UDIBw5CHI2EbHNKcyVc41Yn1/ixQ29+3GliftxpMD/uNJgffwc0PAEajgANf3VI8+PvNPzVwTU+ 4Tii8a8OFs42P+7MMD8+JTXHk/y7ZofBfKWtyPOVcweleUC2LvZ8ZShkfr5Sq1MJm52vdKjmygVV Oggjng55/p2//Pfli9YZ/305u38r/r5cSQeR7Qj/nb9gYt7YkWXemM1rngZl9r/LCXkfyuxid/q8 Mds3zy1dJ9Pfly8ga8lVpBtS92qyCX4rc8YX6vydubqe0Xtv4mMo/JfAxhOhItyj0cP2PenO/t7E YnKZqGcNKFoP4emkExSuJ6dodNXJ6ybNem6WOe1C6umtkHkeAB+HSu9hTrr2Nm4H9z6/gzNTTxcS 97PlUtyHV6Tifr58B3dYxcjH/XKVFPfRtam4X6nawR1Vay5uvWPsqvrHXiLvTK2Wy/XqHL4zotQ/ Vr23Yqb+0epUwmbrHzV7Z4m8MzSsDZbr/r0/Mq77V1v8bZHVRWiDnao2mNPUTepvv+i1CfYsbQL7 qPqPxhHyR1ifOZ6QM2zpdTDbdxxscGV5l+gs8Ysjl0HKszp4LazJsO+ONGvq3WaNHq22O8GvhfjP gw1LQdvTGm1s30OwoTxL+3A6tA2sdbiCdIjqrgB9V2vepauT/UyVxrocvpFCM9SRVKeOZOn2a9jh 5qQ+fDcnpe1v6Rrut7SS4xordetJaqIfTw368dSgH38f7OiEyN+C9Q5OOu4t2gk6dnD3wZqFs/Xj qaYfny8X1nbEYcM/ww0NwbpXbjs+dW+xf+ruKRqT191SmlR4UkwqPN3cp+4d3Ovubo6Fi8FE7164 VB4n2m9BMf3Hye+jdsPFdo6XvmPG3ik8Z9yvhAfHM7+enjOO+V8JZnhOMXhPT/2Os/pd5uAofZ90 tUXjswON+023DYW+T6p9l1n9Xm+wFPtHchs9b5txG618h63QNlpJBzPfYSu0f8Qb/D2BFe+Xs3bw X6E8vwIF81woz0drvgvG9jXw2dvBxTBa6xbf+b0MWkHWDl5NVsH2WeCnye+2TleFD5F9lWpfpvEc G1/+DbTMAi2fVMLYR6OT7buBzz6+ZOO5jaBvkqhQ0qfo0mrKpmcA4rwH4pwI3D7V6GH73jGhJ/29 fEJOVumZo+GWSQ9LAxtw+QrifB/W3xfS9bB93xGk+tdIz2JxPHsWOQ+OmCP2baDfpsNCr4/CZWh3 OIPvYb4GFf5kELQULvCI/D3MJU2r+QubzuTPazqcP6PJxp/Q9Bad3PQ7WtX07/Srxh/Ttxuvoc81 XkJ3NJ5Lb2k8lXY2zqLzG4+hhzVOo183HE3faDiWPtEwl97dcAbtaVhMlzasoqc23EAnN9xDKxt+ Sb+Z8Acan/ARfXfCeH7/hNn86xO+D/46CN/Lvzvht3x8wlv8NxOIUNkwUZjccKJwasOFwtKGK4Se huuFuxv6hScatgtvNPyb8HXDDuGwxp3C/MZ/Fzob7xVuadwm7Gi8WXiu0Su83Xip8FXjWUJV0zHC 5KZxwglNH8G97IV7+g+4t9t5do9G37rUGyPnypb1c3ZCob8ALvIAML5Y/hunR+qX8PfB/ZqJO1Pf QE+DXj/nLoj7BLjY67DeKP+N090TThCZ3zVhI8/C2fo5nEE/Jx8mr06QnpnMggvcI/f9/jRhG3ds 0zZupNNjqFrqjyeqpb4fi/ur6kruq+o1XLHS481qqS/eVJPqdzbVdIKGHdyb1Z0cCxczPdg8g3qM Is0zsPFJ8ZiweQbGhM0zKEyOaOwEDTu4dxo6ORYuBpNM4wptX5yq+uKrIS1vB2EvQOP/epnUJ1hT 81rZc3XMb3WuqWH+tTIzPI3GFupn9EbzJqXQL3/oG6m/y9bF/i5vLv1BrU4lnMu8iXqupCTHRHJf /IUdxn1xds9W9sVFniM8JhIM5q+cmmOmGPSX8plbYWl3fC0hf4FI/gbrM13pfTu2b7ZLKvtGfbuL xH1rxadUa3T641Um+rufQV1zHsRzEtzEzzUa2L7tLrP93ZW6/d3pOfR32T1Uwn3/AuKcCHo8Zel6 2L6vTTA5A3q87N+5spY5Kp9tDMDGTi/AfddC3KtAwzqNBrZveVn2sZMXxiOb4I7Y6Gk1/M9GUF7V mGSaakyi5XSIDiu9/rk9Q1tkN2ift4D+V+GEu2tS7wH8C7QnP6n5k2A0h5KtfbYbtM92g/Z5HcR9 L2j4Jaz/Sx6/rKv5L+EnNfcKv4Q1C2drn+0G7XMuTDK906Btn52q9pl9q5q1z6zfp7TPD45/reyr aua3Oh8cz7x++2w3aJ+17zWo506Co3TeSmkLrGqHzbQFhT6XcWrmDEuyDyS3ww/958jPW+XybLLQ PpDdYA7LluGYbO1wtvks1gZOho1PwMU2QHmu1rSBbJ/dVdxnPkthYyPEuQD09Gj0sH2Xu6x55qON +1/BvwF12z/B9Z+FuPdp4mb7XoJtFRniXi3ew2XkcvFZ3BXJ2bNJ4tsfnWltnPqZnPa54SGaNjFb P+ZC2PguaGsF3adp2my2b3pZ9jSUdK+Qv3Cgr0WtNVMfIgGszoI47wM9D2r0sH0/KZE+RD7f2rkO 7un3cMKPYL1F7kPcUP3Pzhuq/yAU61s7l0Dc20HD/bB+U+5DXFL9pnBD9XbhflizcCHf2jGaIxmt 3z6xqi1W3mMoxriM18wT+ku4La7848jPT+ViJ6jQtthm8N6Iw+C9kWzfoGHvjfwZ+vsUKr21cAFv eXqdyPb9oDz7eyOp8S57a88rpvdcg2/QKNvqstTXTFsU4r8e4r8WDn5fo43t+1PO2taLbXkmbU2a d0XMfH/JUarvUpWovSwr3mcu9FnPk+A3VBIyD/LQH6G92qXJX2zfz8ql84zyVzeEl0HPxAvpu0x8 l7YD+pkryCqxd7AM/CpyFaS6Ot83a95Znaljy0n9TTDl3dc6E/2GfN7T2dAkzV89fSj0ceX5q+ua lvBPHrqEL9Z7Orsg7tlwsY9hvUGev9p16Gz+Y9Cw69ANPAvn+p5O8WyomRurT2mRxsBsXcplU6tT CVtpQ41qyqZR/ZrL3yz8I82BFLPPYeZvFgQT/Yx7J0rfups1afi37tg+9q27bG35d8Txq9KWs3ad iH8fUKd5B7Qpw98vZGvLOawvilpfcAZzOx7N89l8bDGybyzWN0j5rrppeL5j+8x8Y3GBON5nz0qu hVZ+OvhN4t/L1Gna7bmaPmSmvMee19TIcVXK4f8DIVs8yIDtAgD= ------=_NextPart_01C52880.DC295B70 Content-Location: file:///C:/C9344E54/Paper14_files/image005.gif Content-Transfer-Encoding: base64 Content-Type: image/gif R0lGODlhQwKLA3cAMSH+GlNvZnR3YXJlOiBNaWNyb3NvZnQgT2ZmaWNlACH5BAEAAAAALAIAAwAE AmoDgQAAAAAAAP///wECAwL/jI+py+0Po5y02nuA3rz7D4biSJbmiabqyrbuC7+BQNf2jef6zvf+ DwwKh8SiIBBLKpfMpvMJjS5nxqr1is1qtzSk9AsOi8fksonKTavX7DbOa47L5/R6He3O6/d8HdwO GCg4SDiC14eYqJj1V+j4CBkZdbhYaXnpJ6m5ydlZQokZKprY6Gl6iioImmOghSCUANT6s8r2GhRL W1tTmur7Cyy1y5tGNXyDdtxVOWwMS0zr0xtMXW2tcnzbOnvQte0HfTQrzg3d7Z1M7sf9vY137k4+ jhx+Lr+Mb69+tP92/Q8wYAhlh8al4zcD1MF0C+ERW4hvB0OEzvjle9cD4rKG/+XyPYzoT6DIkdYI 0qu30eJJixNXJvx4MdvHlx5rhgNnk6aNliwfzgtJMqhQUyZ3GvWo8+hBhG9gMq3ZbKbKij2PqlSK khJPqjRrTRsKNmygbPDslc3FS1+/V2jNsuuHbMJOdvPUoitosFvbuvHsdhELODAgZaOYFbYlOLFi MoQPI/Lr2MrXxZQry4iMObMry5w7X9YMOjQsz6RLnxGNOnUm06xbc2isOraoya5rL4YtOzcz27w9 49YNvA/t3sSH/g6O3M3w4sxFHk8OvVjz6WKfR7+OZTn17cGsY/9eRDv38ai8gz8vi7z6f+bRu5e4 Pj619u/r75SP39fxn9lv6f9ayYp/68hS1mN34YLAGcMlOJB4HzBowBMO5pfKc/Q1BRIPvxHWWEUX ymIOLhtot5wXk03ogYkaoHgGhSSRVddc75y1izEC6pPMT/7lKE5aLOXVY1UJ2RgkjmdlZJWNOXIz 4ooRjvgkEhmAYGKUf1z5JAAMPrhil1ZqWSWWVjaCpYslIQnSUlRlKNFNES2VJko+PvXNm1NdhKed V93ZZlZOKalil1o2GSiVXjaJ6KCHkhBolE4mWqigikpqJntoYnggOmlFxeY9GWKE6Z0M3SijnHjl BWiNSYoKlJSIwiFlllQ6SimsEMqaYpawUgqmrmPKymKlmnCKKZyfamiqnDf/KbQqT0YZq5Gez4IY IqvIEPoqtpNyOWiVryWa4kAd7Lotud9OE6ywkcCYqjzu0OWVNqgy4O6mQ+r1br3y9ujWvfKy1S+n N87rLpRigtnrowZ7GWuvvzr8mq0SOzylAhA7qeu46ao7ln0e7znExsZxbOnHJodHcsqDnczyaCq/ LMeHLWMnMsw2ryDzzNHVfHPPp+kMNCs+Dz1J0EbfR3TSTOR8tG48Kw21lk0f/XTUSTM9dWxVWz00 1lmntjXXPXv9tWhhi20z2WWDdjbaL6u9dmZtu50y3E6BLE1BRBAEG49xRYarEq7SbabdKeEkjXBV MItLeHtLuC3h+BlO57xL/7rEF3+aehMkv5vze++zoH4el5FEvqt5PH996avEYZ4ruYuUs2KVtHuq ieycXenk7IzSGusmn7xn5GGjV2oLaewczx6qVji6RK2doz5fO+53reK8WbXTnk+2Go+JceDK58c8 gFttDzzi1j5Fz+h8vt9nVdPabqr3ycM+PsnlX7/p9f7+n6k3ACx0sdgX6gBYL08FaEZ9udwClUSx 1i3AYHPLH2f2FzcDXcuCscNgBveglgpyUD4e/OAiRDjC9ZTQhKRIIeEMV8B/5ctelBNQ3iAjwBvW qG/f85YLbwZD9mGkKxia3ZoS1ynuoUlVIhqXoFD4w9YYEUBxah8X0LKqz/9pwxyo8olXNhej/mWq FI1CGMLAR0GMRZE4U9zeVbKnQfkNz1Pnkx/35vg7kJHxica7H5nAtUYpKs6NoHJWf/wSraRI70+k W6DtlkQXXuTKXJG6WMIoFkjbtDF4hawWCPOoFUZqhIm7y+Idw/WtJ6Yyct3KJBv5QCMH5gKLygkh wfYiy9B1bh9DAh2NFDilSyZsghG05JZcaZoVslA4yISaMpcJwmYq7ZnQzAMUpckbalazDdfEZm20 uc01dNObggwnesZJzmSa85zpHNs6z4POdvrmMc3wjky+07d3KEwM8ZQnIcrHxDgGBzf6dBCKetFP f6oClrtjoIWAFEn+Oa3/dEqq3jEbFMyMxioBToSSQoUSRM9VUTJzeWRy3Pe+gqLgj5IaHLo+CpiQ UoR6JFUWSnNzKom+kZUiINMxzdVRmIJFpooM3t5MOTpwPoN2jFvdSpGHP6AmVKhzqOEAIWmEGP6v XwO1l1dnuSXxLSx8j0Ijr8RK1fm8E3CSyRUYpprWOCh1rdlBZVxVCDYIMEKv4IHrXY1DV5r9tXCB vY5fB/uiwu4MsRSaq2JRxtjJPfakkZXsZAdaWRLmVXPrwOHfiMdZzi2uQGoI1oIgd4LAodVQKTho UFPL0p6ucjyO7VOHjKrE3AU0qza54lun0NrXphYbxBXXJ4rLq+3UFnEV/x0jFxm4kdA2N4xaxKGH itTL6kb0CNhg3cQwKYOItW62F2WUR82aXPOqMXzfhRSwghm5w54JbKJ8LlcWibcQwQl3RSHSj5oF Pe7izKPlAqQLgNpSQNbMpQYOr4LBFVsEC5c6y1Xi8EJ5ODnCxZHwg2RzO5ssioJRGyyg5CqrdjD2 PhgGrgvb8Xo420mRSD0VLiJ+AZwTO9rYv/Rzo5uiRb+imvZ+CU7vU1GJ0AkTt5LATZ5Ugzpj8tRY RtA9EC5LNcPMRdeX8FJg57i6ABriEoKwtVgGOCrMA5M1jWj+qXgi9LAGlXm8xeRom+Gr5OlM+bIa IsrSoLhaVeCVz067zf9vhbpnQguNgg6IQQQyWyFF4xTSgzasQupJiv3VhdI07muockfFK4J6qZcS MKe5k2ivwuVU8foRAhtJ5S92VsugS6D65Hvq7qhGbz0m5YWLVVP16QmPo8Z1rn+R6hFrL9RZZMsD Zy1rYdv6prk19rEjjVO8YZi567vv4kB9RCHidqfX1vOuAzZt0opZq2L+toYy18suxs/U5S5OsuGp aX3Wmzn3ljS9951Nf2sN4PYW+K4J/kqDo8baCB+Wwhfe8IA/PDQMjzgk+i3pilvcERhXtMY3/s+J Uxzk5RS53EjOmo4T+uMoX5nJT95y0qiczyyPOR1mftma2zxm4ZEZvh7/Kl28IAdQCMKznA1h9Igd uLxqbrmFihHu6O02ifTtbUYazC1DGHnrLeKp17XudB1a1KEB/Ox068RLDrNp5knVrjFe9V4VJZ2P PqyVBMXlLTjf6ox6r2SgOU2stltdJi+BSOHvlkdmV/3Gc4SQ/Rhs3MEVOMEl8tWilE5knY+PWObj KqylQkemFFCAy/bx4rutR+QVaqNIX/2Jb3UiGUM1uZL/OsE5j9RPY49zhmffsHNLdbNhZX2SvLzr mZxKl774618po/FpP3mSs4t/CqAjZPDlLnr9i8oJpOWuy45uNKf5zKw343ob1veHuZnRw+z7miHs VoTjPGv2pMzypf9y/5gLZu74zz/gdt4Z8/dYmgeAjOF//1eAlSGAikWACRgGC1hYDeiAXwCBgSWB E1g0B3gYF4iBEqKBG9iBiVGBdMWBIbg0H1gYJWiCgoOCo6CCK+hoLTgbMFgdsJRfOYSDjqEjZFFa omU+6xJ78VdiDyJhQ4h1RxdjgtN8n2BQBIZ5ZRBbDGODUad4y7Vto9aDV/cI6RJl6tUEL3Vc/MRi LZBkR2aGMfZHoVUMM0Vr2bUsQYddFYVInrd27zYVBGNl2aU926V350VGDXN+cDZWs+I6hKgxikJ+ ZnRnPuV+g9iHxnVJ4jd+4iVezdeH78VTGmV+ghiIfHc8wPI9hEhnl//4U3bVLRi0bXqYewPye52S SJxkW/eVij5YD97mDYsiVo5CLq4ieWAoe2DHR7JHK2WIhpl3WidmVxJmYlkHiH3EWgqjfEVWe2Xy i0Vmir/Yi1nXYKj4bPtFQ/PWe031jXXIbd1nEJ+WeOTWSka2ixXTSpKIjFxXK4SSILuojWfFfvKI j+kVVsnIhJyIiEe4fNkIXvEVj+XCdPHVj9t4jStyQp5EbOnTPPqlbeioR/FDhTpWkeloagTpj/bI i15YhK+1KxFmjXt0kEl4j8u4jCn5jAQ5ktEHk+Fyfy1pe5RHk0R2jw55QpYzZkT3bKRTVKp2D2B1 fTLkZVe1brFEDu3/Jz7kJwGWNElpRIQaBZX1iEbih5UVw5V/FyZm9QAqJpWVuJViSZZOCV5Y6Ynp 95WFaIhk9YiTqJJSI4OYsISV8oIyRyV1aZdTSYMAMYJrlZd/WWJ8eQmDSZjYAEJX1S6mo4ZX5Fmk 95hz8m5qOJlIk4+JqVZ6cHhxQkQn0UZ9M26KB5pYqIXJ94yayQmbRI516HM29Dt4yBFiNFOxyGNv JJs5yXdpiX6NqJo8Z4OEREiXuQWvyBGyWC2k1GF+skGUCI2sBIr6+JvCMEiwCJE32HOIdJ2zWEeR 6RAjZZTFhz/C2JC1N51mwJoXeZ3YuYYppZ7iNkq69V/KUkRIhpOq/+REiIlyVtUuygYv/JlTeUiH BaJuaddAv7QjEsR6UPl+0nmeGWiYJwQF+rmfEbobX/h3DwqhFvoYGgqYHKoIFOqhKQKiLTSi81Wi sHSiKJqinLmi1RCY7ySiL0qXLeqiNAoMMbpOM/qiOmpOPLqiPhpOQHqiQrpNRDqiRlpNSOqhSgpN TKqhTrpMUPqgUspCVHqeVmpCWDqdWvpBXPqbzPOYxCltFjl0sKBvGdpkOMpipGCanAkdBHUfTRhc DcmmOMNQrsZ92ekjfAibp2d97gOPothebIkr1HinqfVJIvVG2VFSy3mmP0af5gmMJxmQO5moK7Wo bEhtIbOKwbdwpf+jUyo1XGepk3OZqZ+wqUPJntFgU6akNeAgjpQqW5cXVfCXqni6mD+HVdmZL/Ty eYuXoAd0PU5YlVT5iPDlU7mqqjbqqlklhJPArDiDAdVqrdeKrdmqrdtaXmA6rRQIAY4Wrt9KruVq rueKrumqruvKru3qru/qG/w3SWrqlfKqpoRqB7QxNxOiWsgHr3X6ZkeYZx0VhGTIjlsoOMg1j//a dIh4d9gYdwxzl1xJXg/LJXm3UZZnZ/cairspTINap8h6qJUosAwLjNnYjrRyqesYhB5ZjbTql4qY iTY5YM+5sGO4YtG3jyaLXCj7kY+XZsdajTIGe7JlYsRoq0b4sqj/GrKBqKzC5a3yh5/riIz2iIu1 Go2niqrOR7UIBrNmaJMxGYa6eY1RyLMKErFwGVYR24hp25u8mX5PuLELarEI2VoTxACDaLDvB3tx FrVnOxj6sTSAKyzySrh/+WgslriHy7iN211EiHSMYYmGe4hMC7JP2HqAKIZrOitni5JeyLQOVqsO 6lpjq15fi1qlO7Alm6hFq2D8x4lXmWJViVbHWJB41l5SibFK9zoeG5Cam6wWa7ht6YgZi7uV+q0m CX2R57NQ1rCQaJA3a6nuhZ8AKXeoqbIZk7RD648HCXl2Grqt+yvKOLpfObO5mJA3qbMNqrzdu7LG +7qud1ZdmZnM/9hDA5mamEqutPq59luSQLu3o0ue0su9mXef4ymTuBq9Xfi/uvlkApy8Vgm/a3ao jSbBHLu6Wjm+8LuIbcmgmKSWWel+fWsxVAmJElxWvauPf5upGJxNg9tTLOy4M0zDNWzDN4zDOazD O8zDPezDPwzEQSzEQ0zERWzER4zEheDCjIK6TpCQFTwWDZq5JLIg9roxl6u0muliZfWAAgm+6Gk/ Itl1YFewRuiLZ9iBpai3fZuakkix6EWKyhid4dvFfvSwKeuc5cctFAthdZufKcybnhiJlbtzD4yN qztWn1t3Vmu9O0u++Vq1qkfBQOuvqmTIDdx6BEapcyy2NnfGa/+bu3jXvYx4fI5nqtAZuE42ewP8 vkHrl5fsuoZytOW5vg7oi4+MdZgsv/jLyqpMy5DsPZjsy61cyb+7TwXsxcQcxrYqw2PTlYZ6MekL ynn8zH5rlWxGwsDsZoWowQfTtin2th2suy2Gls9MktHaw1TazF9IxzlMuWn1zkksz/NMz/Vsz/eM z/msz/vMz/3sz/8M0AEt0ANN0AVNxE1srAk9GMsauTD8lM1Ru5AbwHKbBAfVyYKGwKfwvdDbzlD4 xR1ds8IgV2MwZKYLwQmrsBSY0vion7y4zQxDkrXbjOCTdNNsYCfSi8KbiGCpiBprK+Z3vy/tu5ro m/erRmq5Xm//KcVJzcWR2Jv0S7K4610CTNR597HA69MffMpVDY+fWBIfGSk0S9FTS4zyi8qslb1n 7bDJaNYyW3dQO8r4a9UH67y6mMsK3MZNvdZUm4RR2I5i3FJ/bXfPt08j2b4bDdIaTbtDS855bcmZ OK+4HNdQrdbeq8Au27GKTHtnRo9zl2Sl7Nl4rY0lacp8fdMx7Y6Azdg6mbJey3xGrbUOWiGUfNmm C5KQrWT967yvXdkGfLOYfbEAbJLC/JLaQtz2W7Jy/bPWmLO9XdzMHbaSzNInDNltDaYX3M1xNq9r zHwo3MHowsbbnLGM9t3l7NnXnMjYHc2wO7J8y95iglCgHLcM/1rN2OzdlFuo2F2WDaC7mUzBM31/ WjzSDbvOWRwzh5aY8TwFS5zeBU4UDG7Qjju34zt+6ftnHIvFhIxROL10Ch7hwf1kiL2yKq2+LsnR aH3SAJu/H87R/btHWwOylHTHcFeorkzjxzy70rzWbAnIWh3QZlvLHo6zykx3xxfMWLvaBxzDP23a QH3R+QzkL3bcHf7QZK3Wzeuw+prW7kXZwa3k0OfgiDthUr7M0rq0mq3LU7vHe83cA8u1M/nHBF3f TkvhbyXV3GzTTq3nibyb4rx+Kkbf2i3bLB5FYU7oxgHhh67oi87oje7ojw7pkS7pk55rgZbotjsf KxXfjNHPZf9cx4KLxiXOunZq6J6sfPBdYj6d1Bl+cTAmlya85HS2lser0Aet1wEe0jJLwH6Ws5IN umdO1/ZM2k8ruks76q1OzRn10WWrxl5X6vu5yoPu2Eku6oPwUkg76FP+vbg+zxMO65oe6CVs45vg lhCT3bVevFP96sdM6TH1Zwje7uQT7xM6rjJQ7/OO7/mu7/vO7/3u7/8O8Oip4KyO7iSLhNQq5Kh1 4hV9t/H37JRWzLv96yi+8Ed2xtaO0mjM7TiMx5NMd3k87lmO4845llH11Fld1OAOyFed8FDb5Qip vTyc5q0tjcaN5Jcc7TFr10L45OVr3Qzf68WY2IA788tNzJb/vuWB/fJx3ssPtmC+Le0tXuc7G/We a+VoTtu2d71Onsxrzsqu/fTT2+ZNm9HNXe2EK9Tmnuftt+fn1+fhrazkHO6wjqhoO+Frr7TyLZal HfD/BOoZ3/djkeiB7zOLS+WNRviJX/ibHsOKXzjLzpDsvNKOL/hw7OqQ+zpuLNNGp/m7S/CUT9IS r9sEi+WxbdKH3POgL7nEu7zPXY+LTbvYfrtDr/piKPsFi+VtDbZgbvq1T9L3jcJAHUFzS96pjupx 2fK+r/zLz/zN7/zPD/3RL/3TT/3Vb/3Xj/3Zr/3bz/1KeOFxK0VLveFUzMSrdeke/vB0s8WQl/6u X6qiXiJc/ycyEU/7E9jsFS7oprh6Izz4s23HbkkAwQ1iD1nVazOtmhCVFs8MHcOKKk/y0FRd2dZ9 4Vie6dq+X6zTm5KCdx8eD1c0FoOZYS/kAy1BqhLxExXlhExq59j1fsFh8djFZfZGIbTZpNlCrmT5 MQnXoqBP7SmVplo51FbSrOz+5hATFRcZY9gKq84k+370lOwaMwcrneIqMS0xH7MuPkdPsy49RzVb XV9hWwJJGDhq/Qj7HvwEQWNbL04IA23X3CJIHI7hRJpxb4V1dntX8Vh/sbO1t7m7ba69XcHDycvN z9HT1dfZ293f4ePl5+nr7e/x8/X3+fv9/wEGFDiQYEGDB/8RJlS4kGFDhw8hRpQ4kWJFixcxZtS4 kWNHjx9BhhQ5kmRJkydRplS5kmVLly9hzvFEg4+ccUjM3JRFk2cZnTl+mmBTc5BOopdi8pvZs15Q PF6cYgnD6tqjn1c3RU2ablU0adOO3jI2IpcxZEcFiRU7KdesZMV6vW3WZo3VXXPrUnNWy+wkocPA ctmBFm8yNLQML926j0ieUlGEmoIk5TBkyciYVckpqZNlznsifdqpDPIb0oYmexZdaNrn0JWrPYm9 eCBcxJtfh16i+LLqJpon41ZNjDLwQ8s2AXed+kqwOpT+zgZU9nPjM7xp+xOu+LjuO9iV+7aONHX4 6mzP56b/W5z58iDN0ScHJb37+tjWtWaPN8vWXrk529rAN1qIIao/xBCES8FnyNqAP2gMi66oAxnU C5r+CqTOLAwBq9CaDt1oTb8Rp1qPRFjyO1FF5DRcMZEWXYxRxhlpK2bBBgeDEQy3gNJRA5/Qwuwb H2ksUhH4IkEStxRncEKqAU1Mjiok1DPSyiOjLO0pJqMbijMHRQymL2ekvEu0DlGZ7r8r2fxCsKf8 ivDF6+4QghfyfJhyvt7A48S4ONsMtIY3TeyMyxC9TMWXP/GMS5fOvnyGhWG+g1LQS0eLM4nx5gzO M8cgcRLOOIQBtMo8Kz0UUyMZ/KoJ51QFUjAw93I1LTKF/0zsrFp/4xFMO/NYVdhhiS3W2GORTVbZ ZZlt1tlnoY1W2mmprdbaa7HNVtttuX0nyFyZko+eq4Yax0CgViujW4z0ZCRWcZqEEwgcUnx3XXz4 kpNHGQDDDE17F5n1rucQVHfgwxyUrZSEBazs3oQWjo/eH9uDh2DT0p0UUtJIrQOs0yx9uKAhfDXC Y4ndEfhVTfldjjWJQTVV5IOCzThe93pTx6rh5H3yj5jDo2/mh2xEc0gcCQQx5bx6dctpu1pds2gK L/Tr26GxbtlNgCeuMuuvs+F60KvB5ohD2+g9m8iy2W7b7bfhjlvuuemu2+678c5b770JBjdckMXW pN6dff+See+MwDHqyYAZ7/qGwA+vra+zxsLrxhDj6pfsV30181dlILR55S1tXQv0xArbNXKi7eMY tFBW2+5LrVm2U7KIk0w7d9dlm92SzVcPiNBQiVM0dnltc+rOywDMl5m2NNbwZ9cw5hTy4NkZnlPl MH6tu/zuo/P2Ontm6mRFqzcPe4VgHR1D1NtP8MGBB5vQ1ltprbrgwf9FOun3PbS+hVxvDMATQ5p0 JsBiEVAKBiTH2hQYQQlOkIJt21f0Nne1C1LsaAxMXAH5BaAQNhBXYyNM/Q6YMQdmSUhBYaBDRDUp w4nOT+IalNcE56abzRA6DcShDL02vB0tDojqchxMGkP/K+55RQmEudWZlFgcV/mPJp4r3QZ3KD8h YpE9utAMhO7XvChx7X6nQ92uJIW5suCihUkBVRLDxCg9CY1RozEUDzMlovI4TnvHy+KWQgGf8OVM iFAplBl7tyjvHdJ2NETJG5EXqj1UJY6ROtiEztMnn8HMkBsqmSN7SImu8KU5/OmSH4d4pknGMVGo JNAiY4Iq8qQnkLAbVSLVJzpZooZJ3eveDQEZypCVjy6sRE3rCIfHv51Mdr+0TN+E45LOfYVXmVOL mTakuhv5x2F3ohQKgQDAC4nJZL8x5zY5KMxj0I9D+4vi1FZoxORliEz+m+c48+I0L1ZwHvb6Vjwt 1g4R//IzIi/MjkHfpjYIykKhACXoQyEaURI570gwigYRcyhRGT3Gg4YrZBdZmEKNxkhgnzOd5sgp Rld2kZ5htBzDyPnDkbqxYdRjJu8WNT1QHqePp4FCJYc5062Ianu9uk7x6nIxI1Kjdbohy1EpJdQT WU+S4zPey4T5wVs2and74plUJxqgK+aPm1pkGjyrYk4LnRWeZwXrtBCqzHKC9K3KWug67lpXve51 RnyIK1+D91HADpaYfyWs3qJ5WMV+arGNZaxjFStYyOrVr5O17GUxm1nNbpaznfXsZ0EbWtGOlrSl Ne1pUZta1a6Wta117Wvh6s+zFY5mapXVTTSoI4e+Ev9t+1RnRP15zJAZNmzi2yRGodOuG5Jqhgis 4CetSb8ibgGmodMO894J1WWUcFTYTFIUWard09kzMuvTaVWVSc30OI+4Ga3jfZZ3RyhNgZOzLGKC 6gTU9mYrme4sYw+pih+B3NS+WK2lj+JrivllysCzmcl+sZVMnoY0aO8tsFJqWLMGyxGIHcsZhbXU qDf0LYLzGyc3ufgfJrqPIM5hEY4+p4bQrXgu+LxNgLb7NGqe12GwlWtTREJiHyMHbnkd8pGRnGQl L5nJTXbyk6EcZSlPmcpVtvKVsZxlLZs3qEUBsSP+hlxrbE135UVXOLPyZE2KeV5dIKNI6UrMLs8Z FRD/BuwoQVdj7XllijpeZ+pK0yI28lkvvCXyKfzcRCjWOUcII6+dI3tURYcYZwZuambE02Z/8amb onBZ+ZJYYfW05qfSubIpueqd3CD10qYmMoDr255R1sd7Cf6ioI0bPkhbNsAo2yPQPEVrj7osfZXy lHwGeezpWliRVaaadFeGzqRF6HJUxBwJp3ZOJYpztiTk3DZd/KEVv3LOU5ZshLf8qMnxd7fpdve7 4T2ulkabzy9Msbhxmzjd8s/GXo73cmVtS4GTIYbLZnMN4wxm5uJw17D9KDTLzVDpRcqtwBqTnIrp 3UCrV1xwRNia/r1UWNYaob0Glnx/1+GKWfq+qBbF/8ZCft/Hoi/iGGwlymkJX0rCfMeO6hkk6xhz mUcTUgbdGHfIN8g54nKPw855sIUu8dHBj50V9eu2E31GNmbzmhXqXI6VRu6aR11aDSf7VNt9drWv ne1td/vb4R53uc+d7nW3+93xnne9753vfff73wEfeMEPviP/rJfI+znCMYM5lfBWLpb288exy/Tg VM6XWLWCUgulHRgS0pfngTR1qV070FbGHaUfRzEeCxSVE95hMz0tZ2evUsi0U/bksaEysSf2uIgS LvS2rGHKt5zYPy6uKNWpPOq1AenDd3JbbRtCpLXT51xh2o2NBrJT6ktt+zMz4a9L5kyYHfyRL/9U Gv9Kh/Sfn/3td//74R9/+c+f/uoHXm/9jUGNpQ3G+m58wufKEQaqEeLn3oBJh0BJWAqOPXBLKh4v XhZuuF4hVpQvAUHo95xP8r7hy4bF485JvNIC4zhB4xot+hhJYX4kQ44mzzavlzxvjUgvejZOdYau mGTMuqRu9w6G41RtWYBOYYouv1TuvIQvlFwuT1zt9XjOzVpvpXYiAi2nBv0E2HqPe4RrjJrlB5+O NQxvCd3nn6yqlprt9YqPDv5C9M4N+dRL2AopwZCE8Z7JlDqmDbOQ5mDPDoMIKUrlVIiPqwSMSkQt l2zvB/qLzmLPq0CKUKrhDTVNwxgR9XCPRqYpn7L/Ti7OiOv0zOtwzJPCjtzAKZye7Z4CUMcKsF8e paXwD3lMitri0H5iChXRJg1di/w0sMxST9PczLBkcbWMDMP4r/Pcr6EORRhpsf6M8RiRMRmVcRmZ sRmd8RmhMRqlcRqpsRqt8RqxMRu1cRu5sRu98RvBMRzFERw5z/cKRiLCrUcySN+CZLfu6V2KseyG 6A3j8fgGjgFxUUpkDgLFkAkVCLq4D9rwMTJ2cGEAzR7OJ/s8MAQt6onYAryMkCBTysUqq8fwhgiN i/K+Z8QkLRLDgcC86gg5TB8x8sIi0swCzEnq0Vj6K+sQTLxAEuYQMgybLyQnKQeJcMH68KmWL9U8 /xJsJKx2nO72Sk0QE098HDHpRlKKmI7WlGt7xoO9LLBuTKxWUGzfXCrGum4l52oSm6YTG+QSGTIT pevrJieAcOxArIYrn+sf2DKBZO/IepFb5nIc7fIu8TIv9XIv+bIv/fIvATMwBXMwCbMws2bNGIoD K++3GCz/1M8WLbLz3nKCELMxa/EAEfD/4hJQIK0yTwvPtAnQysjPRLFywCVHoIYF2UoaDO0sXfHi 9rD6HI0GwzI0yWfJQg3TlMoKTXLEFrABcUrZBAQKh3KJApFwEobZ9AjKRHL1NpKiOM0pwdA8ik0V ru8BRfLWwqIMbTIDZzHXVi6XgI0jf6hdSjLgjv9u2ajKKDltvWTy+aCPFKupKkuo2nDw+vDNntII +2CRi1pFLaXtFNVK8wxQzRRTAW8x+GIMWeqyLg3zQSE0QiV0Qim0Qi30QjE0QzV0Qzm0tHSS+4Dw CxkGMmNQ4u6PHRHNZCyqfgBGEe/zMk8xHYmIbJxoKllFKXenYVAlKg9PHxsRDlUOAH8U2bTPDBXM Ow+UNyVPq77K+CSx1VSIFEbuNYWUutaQ4piqIcVS984xXWDQqVoIIgsLNvBHTWwU4QZxBpMKIJGU pKD00CxuONrNPC1Mn/wwDxERJG2Ix0xxEYsTkHQNERPUQQOx6RhTUHrKlsLoHkfNPw2uZkwOBWX/ kCg/FNR8zQ5ZbR8JMUR5qUAt0wXRML1YEkf14Jfoi+YSNLloSUmVbgh9Spfw8TxxLqjmKFA9c6mE rQrbU0gDJeygj9o+1FNtqET7R6w8SSx15cTWcBPXit7Ay8Qs8wb/JxT3r+WuTgDRcj+TbzKTpRwb ggLHj0u89Yj85v0IdQKNlBu4tUPZFRqFNRV9yEQds4McEzrTVTtBVfrOlLh2scSGL1cTcAEjkQ07 MjP/JF/zsVw1NU3zcV3r8NbyCUtZUUu3Lg55MEu+tBW1UrcsVcUKI2Mx6lYtdiLRaGPbKEmxJ1LL KmYEtlS5c80SVU03jQo50DQ+pk3xC/dY9lTd/4NJISpSWZV4dq5QLycRL/XAMnVYfy7YhDWS4HVG m3P7hslhfbDCeopl8dTDukr2xrMncyphl5Y8ARES/+gOm81cJGpfinXcBo0EA+N/KPGixiR//Ivj 7o0qYrNuq9XbzpJcyhKM3oljzbWTgDQq0m5cbcJcEbdH1m0xHbdd3ZVqIfdeJxcgJLdyyRVz++Fy NRczO9cXP1cpODd0Sbd0Tfd0UTd1VXd1Wbd1Xfd1YTd2ZXd2aQr7fOpfCDCt4LRs+3WMevdMCTde Tev0niMI3UXknAnxninMftdJUXanUCtRk01kTdS54qz2hrV5pzTiSuqJ+k3zyjO1ouZOZTNxw/+M cv2RcOHLvqjrfD00OBaUem1OaaMVfe2XXBXkNnfPc0cLa6+KUS8QSDUzfQl45vx0M+kXtFQwFOuN AJvEP/22HY0ixQwIWqcvcB1odFULcRdXYUmUKzQ43cTmzWjXQruvcn7lXEGx3+YVQwUJPNMTS4iz fi/UZklVfqn0JCU1z15UQnmwa8fv2njUUCPUhn1StrRUOekkhAnPiA8RZ4/If13WhVkYxs7QXaou upazhM+BiVlXhbk4jMV4jMm4jM34jNE4jdV4jdm4jd34jeE4juV4jum4ju34jvE4j/kGiQ1P/MQu nYq0cd7VU3NLgvmvj6Gi5xbv+xBYHkuEhvn/aAxbKYj/l0gdt2XFVXnT9RFN5QGfBSAxzwRHMFd0 suvM1GpYhJSv2HyXyG5BMFlrtGJlloJnU9tY00Z2dy2nYIvy1oM7MNaIqhFV0qY2lY4qDaiImIB/ MNxMbulklVZVr9LaaIaXtzqGuVA/maV8dTpXrqiUBPgeUr/MkeBo0oDRi7vUrZsrVYer8z2m4xC6 NyejKlqCcnvptKu8GQNROWi3ViYyTH83DKucGXBk5imleXqoeYtItem8uCKqctqgFdtU0Ky0qPuK VqGmrmk38ByNpjSlVZZ5YSspklmzzYqBVVmndT9P2BOXJzLdrqFZR49leqZpuqZt+qZxOqd1/3qn ebqnffqn+WtgtZdXDzWBARknPhgMYfqwcBh4nZqo+fd+wxeBm1rLQPOaTrNvU7gSw7IiRVnrULPQ cNk1JRo2i7dUltqxcjMK9/mcM804flOY8dl3jil525qjpFnvorafnxM7xBbRdLebi2/W/jRn4ReM 401lmxCZgDmZ8VSwqfMQCZZ8b4/vqLVu5/M/T7og0fmrp5XqNg9uMZpvRztAAW+of7oTgXq1VdRD Eq3/Ojip8PNzX9gK17NlqYSaS/d7KHuXqiiJNbbPmoZCebs7U46VPyUmN82HYbKSFtKEatTi+ohm C7PYpheKbS+f9TAJB3Obd7kEERuTOLEES/82tms6rVE3vFl7vdm7vd37veE7vuV7vum7vu37vvE7 v/V7v/m7v/37vwE8wB+Jj+ERvUfwRCGoJgx8pFoUuxvZ/AD40j713UCZ0Lp0U2lw9Eznuxo6IbMr TscSnhd8dZ65mnMwTzt1rh0898oZjoz3iUfcX43Qu2XoYp6KHhVMveHl6czWULw6y+p55vhwil8G xz/tKFMhKXsbdmK8xKLmsiG4q2FltFnavCnZK/lzNE05uZexyaHan6NRxwV8zMm8zM38zNE8zdV8 zdm8zd18/gjIyi1yGAuoXBB8HQWQkv04Vd2mozployGPKYdUk4vac0suABPqxfxlBzPcYmX/EKu9 eo0kxaPBelIvvJXV6JVD+oQuqbxnEK2O9aIMBJwDMh29PCXIU8DYmtKC8m0hUbmDHK9h7bGvOzt/ Ko/oWkn0V8Jk6VxUXFFJvMI/xsHGWdwCugeT215FrT7sWonLmQuZ1Jfmgyc72WsnedGykm+QT5KM nGyddqEV+tmXvdrZ87iBfclxuzrrWilLvLgjyZ732IpSmvreUaIDFNS/koouuoc3W0C1Wi0DEo3s J7S5Td/JasYE7cnjU8zthoSdF1pO/WHUW7W1heHfvM9LqaMfWmrkPH2fNpAfR5EZ2Zd/Mspk/Y4e zJpX3KjLltDHlpNXDZITO5iaFGe2OIcl/2RBSB5jVdOWHY3aQdbRMc2psDextTL5MqnkhRzk95E1 WT3X9RnZSSayyz3mePbaoc4z4XWenVd2fh2hORXZb4PY7PTsohKboW7ngamoElbaiTbqTfzYSSzi f3Z8T1q2OT6RaczIdkw/O70/8X1ts9iwITfi6f7iET/xFX/xGb/xHb82KMkmZgUY8xwlu9GTH/mp Eb1hnzcZL8/CoZusNR9/tXXjzloaiXf013PpA93Ed3P05Y9kSP18y/7hxxZ/epn3UB+godjrWf/P mc9oO/8Yf1WjhfiEPz6ASVFERVymG/zxtVEY7W+lob/6rf/6sT/7o27UExOLcd63HhPvj/9a7d/9 wSMTkQG9hSPtQLE+eIna4a+Q3Mmw0DuW0NHW5dNWW7vXh8QkFjk9ugkA/kVqqHVbLkKK4EXvRsBc woFamGljJZkPRrKeWoFt6dY2nus73/s/MCgcEou/EWyjXLkmjBAyKRVGSajJbfbaPKXaU+2z9N6s 4PN0+S2fO1rkOt000uv2Oz6v37OctGhVWF/XCU2hIRoOiiDjImHK2FhK3AykDZyZymCc3EqlxeGi TiBfqekpamopIVMk2iNZ7BBpFqOaJAzW7StUretl4q9uMO3XWy/uqOoyc7Pzs8cfhoyEU6gMRUe0 ZVDxZPaHdss3NTZ13+b5thtiZmcSNzn/u/jVdPn8OA8tNH+//z/AVfsCNuPEhmCrJgMRMmzo8CHE iBInUqxo8SLGjBo3cuzo8SPIkCJHkixp8iTKlCpXsmzp8iXMmDJn0qxp8ybOnDp38uzp8yfQoBG5 7VA3amEdbVuMIF1gZZjQqFJ7ND3yS9HBOaoAJe2WqOrUsDOzoTvHLhpaslq90HNTQtobpd3GKeXK NBQ6JpbAiu3rUhwyT0vRdjIW2M8wWPCKHNuV0GsZSsT8Ut5ZD5TgNAZjNRbc2c9dvK74ustrFNPj yqphimq0BXVhx6xufVZCOrKtrFbnBIJ9dTVwlXvpXlk3b5toTeuMK6eLnPARe+bw3Q4D/xcx4t/B t3PvWx1Y9/Dix5Mvbx7idKNE0qs/7/49/Pjy59Ovb/8+/vz69/Pv7/8/gAEKOCCBBRp4IIIJKrgg gw06+CCEEUo4IYUVWnghhhlquCGHHXr4IYghijgiiSWeVIVcKH2HFR0rmpigixfFyMeML/KHCFFU iREDZs9RAUl7uin0yHEvpKgMBzZCaNcsuhzTxWYIPaHYFG4paSKTVFhnAZQ56kjckbPkMogv71zp YZaQLcbFmGY2NKUsrdV4JoLqeFkUOMaZQyZ67O0p3Xpz0jkooYUaeiii1aSXlJ+CKjodo40mOiml lVp6KaaZaropp526iWc7jN1WXCuOUv+lZpCn+LabdqbwVU+LxLSGZKv1DQRVV1qysWowawHxaq+p 5SpkDrOWBp6aPvCq667CfsWdWtGeRR1WlRBGXC47RhJqqb1expxpw1lTV7JspRUPPedaC1iwtG4L abhwlWUlq++KUdeOaS7rLbROMmnlreQmxma+ZRYzGYvd2sZLm3iQOhsZ41pXVLkIu5ZZlGIKAvBg x+6bkKkyGebOyMVqJgIWcDoGLrLbHmUQkNcQqzFtT00DxV76VHzIaPmg3JkdKR+2ssftggztW2uV 3FvNnI25hrEJu6xzxEmzCYajcjrtma9Sn6pzbVWbIajQV6/p5qq3HqsakNeRk5fN04ZY0yi1fEqd JsoipAP3N9AppOzbgFbDZ9th9kzxJfLqybch+xg+5HGC3ylH1C2HLLJFc5I68zOPo/IdHAtdbrIi o68dWeW+mv6X5wFx+xc0r9PMkOyrwy1eAQA7 ------=_NextPart_01C52880.DC295B70 Content-Location: file:///C:/C9344E54/Paper14_files/filelist.xml Content-Transfer-Encoding: quoted-printable Content-Type: text/xml; charset="utf-8" ------=_NextPart_01C52880.DC295B70--