MIME-Version: 1.0 Content-Type: multipart/related; boundary="----=_NextPart_01C43995.4B4B8F90" This document is a Web archive file. If you are seeing this message, this means your browser or editor doesn't support Web archive files. For more information on the Web archive format, go to http://officeupdate.microsoft.com/office/webarchive.htm ------=_NextPart_01C43995.4B4B8F90 Content-Location: file:///C:/0D741EA9/BIBOBCl(correctedt).htm Content-Transfer-Encoding: quoted-printable Content-Type: text/html; charset="us-ascii" Copyright © 11/1/03 by Ziny Flikop

Ziny Flikop

“Good enough” phenomenon and predefined bounded control <= /b>

Copyright © 11/1/03= by Ziny Flikop. All rights reserved.

 

Abstract

The paper is an attempt to generalize an approach presented in [1, 2]. It discusses how a “good enough” phenomeno= n, honed by evolution, together with inherited and learned experience that all live creatures use in their struggle for survival, can be used for a decision-making process. Proposed approach is a variation of a bounded-input bounded-output (BIBO) method currently widely used for the system stability studies, and bounded-control (BC) used for the control of different systems= . The paper proposes a method for a very fast defining a bounding surface as for = the input as for the control subspaces. Enhanced “bounded-input bounded-output bounded-control” (BIBOBC) approach provides also adaptability to the = bounded (soft) control and allows transferring controlled system along a suboptimal trajectory.

 

Key words: decision-making, bounded-input bounded-output bounded-control. <= /span>

 

Introduction

Survivability of all live creatures, including h= umans, is based on two fundamental principles. The first principle is the ability = of a creature to re-use an inherited, adopted, or previously learned experience (solutions) in different suddenly arising situations.  It is a well known “If . . .= Then . . .” approach. The second principle is that instead of relying on t= he “best” solution producing the “best” result for the precisely defined situation, a creature uses a non-precise solution and is satisfied with an outcome that is just “good enough”. This flexibility allows a creature to use a limited set of solutions to cover ab= out all its living situations.

 

Let us examine the second principle on the examp= le of a human, because it is a less obvious one.

 

1. We select a “good enough” dress b= ased on forecast, the type of activity we will be involved, and on our previous experience. This “good enough” dress can be used in spite of so= me variations in weather conditions and our activities.

2. We separate apples from oranges based of our = knowledge of a “good enough” range in which apple parameters can vary.  For example, it can be variations = in the fruit shape, color, weight, skin texture, and so on.

3. A physician uses a database of symptoms and a history of diseases of multiple patients to find a “good enough” range in which health parameters can fluctuate and indicate, with some probability, a disease.  =

4. We recognize, with some probability, somebody’s voice if its audio frequency spectrum is in our subjective “good enough” limits.

 

 

Some examples from psychology<= /i>

 5.= Every human is unconsciously evaluating his/her well-being. This evaluation inclu= des analysis of multiple directional oriented parameters, most of which have fu= zzy values and have subjective importance. They are, for example, health relate= d, financial; family related, represent living conditions, and so on. In addition, everyb= ody has an unconscious fluid in time meaning to which limits these parameters c= an fluctuate and considered as “good enough.” Depending on such ev= aluation, a person is either satisfied with his/her life or not. For example, a person can say either, “I am OK,” or “I am not OK.” In the case of a negative answer (solution), a person becomes motivated to start an activity to improve the situation, i.e., to move life parameters inside of “good enough” limits. If this activity results in improvement of the life parameters, then it creates positive effects on a person’s psyche. It should be mentioned that “good enough” values are ch= anging when values of evaluated parameters are changing, caused by either a person’s activity or by some outside factors. If life conditions are changed and a person changes his/her activities to maintain “a good enough” estimate of his/her life, then the person demonstrates his/he= r adaptive abilities. (This remark is outside of the scope of our discussion, however,= we have to mention, that if all parameters have ”good enough” valu= es for a long time, or if in spite of any activity person cannot improve value= s of the parameters that are important to him/her, then these situations can res= ult in depression.)

6. When we ask for the advice, we unconsciously = expect the advice to be in some “good enough” limits. If advice (solut= ion) we received is within these limits, then there is a high probability that we will accept it. However, when the advice (solution) is outside of these lim= its, we will likely reject it. One can continue this list of similar examples.

 

Now, let us consider how this “good enough” approach can be transferred into a technical decision-making processes. First, we should define a controllable system (object, process, etc.) as a system that is determined by a set of input parameters, a set of output parameters, and some mechanism that allows mapping of valuable inputs into acceptable outputs. A system can be either linear or non-linear.  In the case when some system param= eters are fuzzy, for example, color, taste, etc., they can be defined via some digital directional sequences.

 

We can define a system state as a combination of system conditions, i.e., as a point into a multidimensional conditional inp= ut space. In the case of technical applications, since in the future we will u= se a predefined solution, instead of the term “solution”, we = will use the term “predefined control” or just “cont= rol”. Some controls are relatively simple. For example, dress selection, separati= on of apples from oranges, recognition of somebody’s voice, a patient’s diagnosis, and so on. Some controls are more complicated. T= hey include predefined controls (predefined solutions) as an initiator and a closed-loop control as an executor. For example, our reaction at a red traf= fic light. Some controls can be defined in the multidimensional control space. Controls can consist of predefined multi-steps with closed-loop inclusions.= For example, a pilot’s actions at take-offs and landings. Generally speak= ing, about all of our adult life our activities are based on complex initial predefined control (predefined solution) multi-steps with closed-loop inclu= sions.

 

For the analysis we can use a real object or eit= her its physical or mathematical model. In some cases, human expert participati= on is necessary. In the analysis below, we will assume that we have a mathemat= ical model of an object and specialized software.

 

In the very beginning, for a particular situatio= n (let us name it an “origin” point in the input space) we can find via optimization a “best” possible control that maps the “origin” into the “best” desired output (in the one-dimensional case see Fig. 1.) If for every possible input we will define the “best” control that provides us with the “best” output, then we finish with an infinite number of controls (in the one-dimensional case, see Fig. 2).

 

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

        =             &nb= sp;            =     

        =             &nb= sp; 

<= /o:p>

 

\

However, as system designers, we may define limi= ts in which system output can fluctuate and, in spite of these fluctuations, we c= an accept it as “good enough”. If we define a “good enough” b= ounded output area, then we can have two extreme situations. Namely, for one input= we will have a set of “acceptable” controls (in the one-dimensional case, see Fig 3), or for one “best” control we will have an acceptable input area (in the one-dimensional case, see Fig 4).<= /span>

It is possible to have an intermediate solution,= when instead of one “best” control; we can find a set of controls th= at satisfies “good enough” bounded output conditions. However, the price we pay for this is a shrinking input area (in the one-dimensional cas= e, see Fig 5).

With a f= ixed “best” control, the more input deviates from the “origin” point, the more output departs from its “best= 221; value. It means that with too-wide deviations of input a “goodness= 221; of a found “best” control can deteriorate to the extent that departed input will be mapped outside of a “good enough” area, i.e., control becomes either useless or, in the worst case, counterproducti= ve. Border points in the input space that the “best” control cannot= map into a “good enough” bounded output are the cutoff points for t= his control (in the one-dimensional case, see Fig 4). In the multidimensional c= ase, cutoff points create an outer surface of a subspace in the input space (bou= nded input). The “best” control is capable of mapping any point in t= his subspace into a “good enough” bounded output.

 

Like in real life, usage of a “good enough” approach for technical applications consists of two phases: learning and execution. Let us start with an analysis of the learning phase= in more detail. We will do our analysis of a generic system that is described = by an input, output, and a control, which can all be multidimensional. It is obvious that dimensionality of input space, control space, and output space= can be different.

 

Learning

1. As a first step, for the given input represen= ted by the “origin” point in the multidimensional input space, we will find via optimization a “best” control that maps input into the “best” desired output. This multidimensional optimization proce= ss is complicated; however, the result of it is similar to the case shown in F= ig. 1. Then, if a “good enough” bounded output space that includes = the “best” desired output is already defined, we can identify via reverse mapping a bounded input subspace, which “best” control = maps into bounded output. We will do the reverse mapping in few steps:

2. In the very beginning, we normalize the input= space at the “origin” point by defining a measure in some units for e= ach parameter (dimension). For example, a unit of temperature can be 1/100 of a maximal considered temperature; the taste of a fruit can be 1/100 of maximal taste units, and so on.

3. This normalization allows us to define a directional ray, which originates at the “origin” point and is going in the input space in a random direction. We can achieve this by multiplying each dimensional unit by corresponding to it random number.

4. Now, we can start to move the input away from= the “origin” along a randomly selected ray. As a result, the qualit= y of the output will steadily diminish until we arrive at a border of “good enough” bounded output. The input that the “best” control maps into this border is a cutoff input point. Considering that a ra= y is one-dimensional, we can use a very fast one-dimensional optimization algori= thm to find a cutoff point for the combination: direction, “best” control and a “good enough” bounded output. Coordinates of a fo= und cutoff point are memorized.

5. By using another set of random numbers, we can define a new directional ray that originates in the same “origin̶= 1;, repeat the optimization procedure and memorize a new cutoff point.

6. After we repeat step 5 a number of times, we = will have a set of cutoff points that determine a boundary of the input subspace, any point of which can be mapped by the “best” control inside o= f a “good enough” bounded output.

7. The cutoff’s points allow us preliminar= ily define by some polynomial a multidimensional surface that bounds the found input subspace. We consider that, with exception of systems designed by man (for example, the Stock Exchange), this surface is convex. In addition, the= input subspace restrained by this surface does not contain sub-spaces input point= s in which cannot be mapped by analyzed control into a “good enough”= bounded output. We rely on these hypotheses until otherwise proven. Polynomial description of the surface can be done by using a “Fit” functio= n of [3], or another fitting algorithm. The result of a fitting provides us with information about the polynomial and the value of the least square error corresponding to it. Using this polynomial allows us to speed up process, s= ince now we can continue selection of testing input points that are very close to the bounding surface.

 

These = points can be randomly selected one-by-one and each point becomes the “origin” for the procedure described in steps 3 and 4.  Each additional cutoff point is us= ed to correct the polynomial and recalculate the value of a new least square error. We continue this procedure until this error is stabilized. Then we terminate it and consider that the polynomial is found. For the two-dimensional input, t= his procedure is illustrated by Fig. 6. After the polynomial is found, we memor= ize a trio: a polynomial that defined the bounded input subspace, the “best” control that maps any input of this subspace into a “good enough” bounded output, and the “good enough” bounded = output itself.

 

The approach described in steps 1-7 has been successfully verified on the model of a telecommunications network [1, 2]. = This model represents a system with 31-dimensional input, 43-dimensional control= and one-dimensional output. For the polynomial fitting was used a “Fit= 221; function from [3].

 

8. Now, we can try to put some flexibility in the already found “best” control. To do so, we can, for the same “best” output and “good enough” bounded output, cho= se an input point located inside of a bounded input subspace defined by the polynomial. We will consider it as a new “origin” point. If we repeat steps 3 through 7, we can find another “best” control and another bounded input subspace that is intersecting with the previously fou= nd input subspace. As a result, all input points that belong to the intersecti= on will be mapped into the “good enough” bounded output by either = of the controls. We can repeat step 8 a number of times and define such contro= l subspace that any control in it will map any input point belonging to the intersected input subspace into a “good enough” bounded output.<= /span>

9. We also can repeat steps 1-7 for the same “best” output and “good enough” bounded output, how= ever now we will choose input points that are outside of intersected input subsp= ace, however are in the close proximity to it. If bounded input subspaces for th= ese points are intersecting with the previously found intersection of input sub= spaces, then we can accept the newly found controls as valid controls for the “good enough” bounded output. This will allow us expand a contr= ol subspace found in 8.

10. We can repeat steps 8-9 as many times as our= time resources allow. As a result, we will have an intersected input subspace all points of which are mapped into “good enough” bounded output by= any control belonging to the found control subspace. We hypothesize that this control subspace is convex and try to determine its bounding surface by some polynomial. We define a bounded input subspace as a not empty intersection of input subspaces and a set of permitted controls via a bo= unded control subspace.

 

Now, we can modify our memorized trio. Since any control that belongs to the bounded control subspace maps any input point of the intersected input subspace into the “good enough” bounded o= utput, we memorize a trio as follows: bounded-input bounded-control boun= ded-output solution.

 

Steps 1-10 can be repeated for completely differ= ent states of the analyzed system (points in the input space). As a result, we decompo= se an input space on controllable subspaces and will have a library of solutio= ns for different operational situations. A set of memorized trios allows us to execute a system control in real time.

 

In addition, we can maintain the same “origin” point, but move the “best” output and “good enough” bounded output along some path. Multiple repetiti= ons of steps 1, 3-10 will allow transfer the analyzed system from one state to = the other state along a “good enough” trajectory.

 

If we maintain the same “best” outpu= t and “good enough” bounded output, but move the “origin” point in so= me direction, then the result of steps 1, 3-10 will let us equip analyzed syst= ems with adaptive abilities.

 

We can reduce an educational time by utilizing an available expertise. For example, we can mimic educational processes that humans and some animals use in transferring their own expertise to the chil= dren. In complex cases, experts can teach, for example, a robot to do different tasks. The experience of a test pilot can be used for the development of a recovery procedure in case of emergencies.

 

Possible applications of the proposed methodolog= y

 

1. It becomes feasible to create fail-proof syst= ems. To do so, we can limit system control by permitted (bounded) control space = and, as a result, prevents some accidents. This approach, for example, allows prevention of airplane crashes, similar to the crash of the Egyptian airlin= er, or the Chernobyl accident.

2. We can prepare in advance a set of flexible solutions for a robot.  The ro= bot will execute these solutions depending on the required task and environment= al conditions. In addition, a robot can adjust its reactions upon changing environmental conditions or output requirements.  This is a very important ability i= n the cases when robot works independently without human intervention.=

3. It is possible to develop in advance a set of recovery procedures for natural disasters.

4. It is a very well known that precise plans developed in advance become useless in real situations. Applying proposed m= ethodology for the development of adaptable plans can be very productive.

5. Systems (objects), for example engines, can be transferred from one state to another via a sub-optimal trajectory.

6. The experience of a test pilot can be used fo= r the development of recovery procedures in case of emergencies.

7. Time for the development of systems with boun= ded-input bonded-control bounded-stability-output can be drastically shortened. =

With some imagination, one can significantly exp= and this list of possible applications.

 

Execution (Predefined Control)

Execution of the proposed methodology is relativ= ely straightforward. Control system monitors input conditions of the controlled system and defines to which input subspace the current input belongs. If in= put belongs to the subspace defined during the learning process, then control corresponding to this subspace is executed.

 

In the cases of recognition, information about o= bject (sound) parameters is used to define an object via a library of solutions. A found solution gives us a probability that recognition is correct.

 

References

1. Z.Flikop, “Input set decomposition and open-loop control in telecommunications networks”, Proceedings of the American Control Conference, Seattle, 1995.

2.  Z.Flikop, “Some problems with the design of self-learning open-loop control systems,” European Journal of Operational Research, Vol. 81, 1995.

3.  S. Wolfram, “Mathematica,” Second Edition, Addison-Wealey, Redwood City, CA, 1991.

 

 

 

 

 

------=_NextPart_01C43995.4B4B8F90 Content-Location: file:///C:/0D741EA9/BIBOBCl(correctedt)_files/image001.emz Content-Transfer-Encoding: base64 Content-Type: image/x-emz H4sIAAAAAAACC81cfXBU1RU/+5FNABEamJYq0U2FilUDCkXFqhBAyPApBJCPLCJJ+TCQkA0sKmqQ D/kQCFqdQabVohbagaLlQ0U0QcaBFtFaFY1ax+kfre0/pVMdp+qM/Z17z31797H73m7YgMv8uO/d d965591z7r3n3nMgQES1pH/f4hcKoKTk79QlRBOB6MhxtxIFqFcF0WOgKSDXLwyEiAbj2d9dj04f ClP9gyECA/rZRURlwKgRFcNB2sTvfH3hhzcx3fcv4hY6gU+XrkQRGqve5pa6BAvxhK8uCH5Mn6qr cKCrPPW+6hTEX8Ei/PVVsABckuJejZurUZYHIAT1A4oEAaGLUvofOor8froFLXUkEBRpguqK2+oT KqZnqAU47GAv7vcBe6kV90cFrynslXr97BVF/6xCC3VX/CJC26LqkngtY99EAvydF6i3w440e1X5 WhZ9q7/LvB/FexqtCql8Ga0KqfStUt+Sk1bbe2XabqT+CnHBWFXfR9U14M5gEc1zUEMPOJhJjwGP K4ynnQ5uof10s8I+hXH0Gwcz8I5BNXgY1IK3wRK0acByGc0yzzJoXeOwAvM331OCPtRoUTD1hl7X t56TPm6/VrbSYGpWGILrIZ60zTSUtiiMxvXotLSm97ZAN5sVJilsERheW6R+E01FOdWz3U3Q+cPA BppLGwGvdtfTAlqnUKuwXmB4rZf6tVRPDwFe7a6mZbQKaKL7aCXg1e4D9CDdr7BK4QG8/QBKw4vv uf4eSHAvvsSr3eXonQR6pQFWG4e1e9HW03aqoyfwxU/SQnrKU8afY2aqwbi/g36rMId+p2B48bX9 zKvdmfR7YA9V0vPAHzzbnUgHgQOwhoOCFxQML31/EBb1IvCSZ7vD6WXgEN2IUXYjvepJOxgjcDBo ylCWZRiJhrYfHQFaqQ/KPlnNxNlc9Qa+Dn4uS1dzgC9CwUBx5YJFNfHo+JpEdFLdojmL6YXXjyVa gWjrnxJEA2QlC8r6xStGN+DazhEaqOoqFtcvbQRVRP0Jn7W0qWtkHDPfMgtx9GEc/dKAsg73jHoB P2sQLHXeOYD6A47+Nd1+VWdQD2S3RsYtKZhTrmtkI92kEBck6/cpxAU2vabVz8/FbGx6vppmYYTO djAT43AmxiNjLk0XzFDQz/TzGoop+mpgLniYnmc65mnAz+5EXXY9b6ThVuZ6vJWp55MtVymY+ufp esENCqn0VVJ/fV569GHMipsx3xushaWuwVrPWI/Vn7GB5ivYzzbRXYp+E/yQjbg2Pcp0zNOAn/Ga k12PGmm4JeaUa4+aVlk6hqnfh3mOsZ+uU7DpmY7r+Hk2s1V3KlSz1WFumALB4I8m3x1vrFnEd52p lDYEzM7luf8W00tHjiX0bBVSs9ZpyOue9ajds16xmvUGYr5nmYbXLW5sqKvFfkHPejz/hSz5f6io u6Fv+fvLa+KN0bn6lSKhZ+qIeiuc87yZOkfeS9fCIxjkIA6bbYDNMpbDi2TcjW9i8DPzfAVmcKZf QT+FLzDQsSumY55cZ7CcrsnSrow03ApzytWumqFXxhbsDhimnqVlMH+GTc90XLdC1qSOmRnbv1tk 8/OCtrawWG3A6tVL1Fd2xZzaHzPSWMy582gyPLgx8MeGYVfBY+0njoefAPpjveiv6M/3V2e7R9Zf HU5jC7PlS6qg11gG3RraGOytSmEYrod5eoExGgW6UapHGVUCw6tK6mfRBFxP8Gx3Fk3BqjQFa1IV yirPdmc4K+c8hRkCw2uG1E/DPHk74NXuNKwMU4FKjLIpgFe7lbCYyRYqBYaXeT4Rq8FtgFe7EzCT jwfG0KPooUc9aStoG7zobTQCu4GRPjuC4bDmcuwK2KqHqeudCoYXX3P9TbQbe989nryGqHX7OVmJ 9slKtN/hxddcPxCe/iB4/F68+mPGL8NOgEeZxit0JWB48T3XX47ycp9dQClGaSlo8rsfT86E5YJh CsmTlnLBMIVU+mFSX+6x2iS9/oFyEpfq9Yfy7PVH0Tv2nyj1ALqj73pgR1SicJmAn5UKLnHe6I36 3o4ONV2JqjO4DMhuRYtaUjCnXFe0UsjFiApSraHFOTmz6TWtfp5bP7Lk7KGUFgRgt69jPklH6+cn 5bo77K6s4yDGVQRXE5Y2wjy0p3M29hFM6cetsNOtYtu2fW/F3MJoFtj0mlY/T9eKPhl2r0lhqy+Z om8oCC5jgMnW/XRgjnU/D6iz7pcC91n3q4CN1v1WYJt1/ySw07rfAxzw0F76PnpZoVmQrG9RaBbY 9JpWP2//7DMbPT1b1t6YpZ3Z6HlGTGDTa1r9PPN3ptNGDNqIWdqIQRsxSxsxaCNmaSMGbcQsbcSg jZiljRi+PmZpIwZtxCxtxKCNWM7amI2eZcQEyfoWhZjApte0+rl3z3fcXkeP5nfgX/BovnXBvOg1 ZYUYxRFBUoZeivIIvIIuZp9TpwY/eq4Af0LqjeQ8wDPTN8HUOEyB4/e6vygkdN+TPmVtdBH5ML0R x6s6ue47y72ZB83zT4GPgZ7Cq2uW173kmjunb1KOpiqJnXE7rPM+sn6NTXrF0Vw8Zq+faZc59JDr oOxJzfVUoBp4FXgPeAv4q9DkK9LEbQySNtqAFunTiGojffQpU7/2sL6J+V5lyX5C+IYt2U1cKhud 9XD1y1WinzYpPxTe7piVLUeblB+myJGMY7n5vnoG32RsK1eby9e1kXEN0B9YKyXL6I6BMe1uYJRF +4Z833ctLmbscJDYCst8SuwwU6zM9EWp6KvU0pc7fmbTviftGFp3TO186TYfdrETuA7YxbF5+T53 FM6Px7N8Gia8xgkPd3TOi8cgNbYIe1DN63bRY6aonWn3aaHld6qkXXckz0/2p4C5wqtWeLgjfH6y /xKIA78CEiJ7psifaXe70PI7K6RddzTQT/ZfAKuAbcBG4eGOEvrJ/ojSN2EXr8GyZ4oemnY3yXtM /4S0644o+sm+GnhS5N4lPNyRRj8edwN7gJXAfuHhjkD6fX8d8BKwDGiV788UmTTt1sm6Vy/vcLvu aKWf7HOAo8AC4I/Cwx3F9JN9BnASmCVrPcueKbpp2p0MvCnv/kXadUc8/WQfAbwjY/194eGOhPrx GCTv3mzNq+4IqR+PK+Tdqy1fwR057Yi5cxLrPqB9i89dx5zNkj+lvG5odxEsKQ5PYDzKBMpJsJ5F 0P7iHHy/smnHEkGgW6dAwtQZn5J/3d8vVuWJEf9LHK97KHRy4xeJ5YU96brmL1Q0hH9/Rj2/X3nl D6SqJ10/4qFQ9ali2oX3dn9QTM/9el0ouqAEDC8NFwFMUx/uSd3HXYz5RPvdQwt7FUSl7SQv/bN9 XZMfxrR/qz+eKJDnp9FeAMzeGFcinl9Pqjyl5R+N7zi97qvEu3g2+s3LqHqZ/sS+afzdgKwXPGcP kQgQrwUXyrP1vbqWFxZ2LR8o46JA/HLXCfW35r0KaKQeO8RGsaOIhXAebSedv94kftY64COhyVfU 28wVK6WNNTJnRNT8lT4Snou/vtKS3cwF7hh5e/11lnWo+KFDHZ81NX5u054Q2hMObWpMPR3fExZf O85+Pv2yalkXeI1YCiwUm8hXPN7YxBLgTvFdqsUmMsXoc7GJBcLXyB62ZDfR+/bahJGVyxqLtx3Z N7QnZW7g8kahdUf7bb6GNpVvMgMg3zp+QnTAvu/9ouN8ZQgYHbN/2QDsEB+HdZwpayAXHW8Xvkb2 sCW7ySdor453iO08LTIb3naugaF9S/YvbwM3CK07/8Dma2htvnZOQi7rf0jOlQqtNeWwnCSrTAV4 qHHMJjW0iNr7y2YtzXb9p5Q1+/yu/99Mybz+h6emX/9NBNrYi58fRnn0w/L9TbYewsIzVQ+az2fv Favy41Oaz7reXyTYZ2HdnrjjHwltmUSf4PrLos47lu/ZEOJTXbceXsT7LPNY8O80+p+Jddd+pvjw eyV7jiX4PZaPSxUzg9yfNB5XJ8TZ+GHLgQ9kzuEz2OEuP+wQrgfKWMnkh81S+4g66KQRM0id8u2K XH6Y8c1CrjFZKfvh1XLee1zObW0ZGiSGeYGHDDwvlytbaYRO5qZIU+SSI2RifZZM4bPwGdP5hqvl LK5Z1jCmyVe2j1kjeA8+ANgs9xG1t06fAZTLGrFG+DZb6687N6i9a8Sz6sxOnzld6pw5peYNGdrN Yntb5PvC6swlNZfIpjV8B1l87fyic+kPVlnnUmcbY/DLQTLjO2zNs4E0up0o56L9RbY2kbU9GUod 2WfnOi7Tz+q/dDZ7l/TZYsu23PlNfmced8lYqpVzmLDq99S8J79zI/bJxwiP22S8Z8qHMu0uFFpu f4q0686R8pN9vpxVLbR8bHfulJ/s80Tu+eI7R9RZVvqcKtNutdDOt/w+d55VNud19wmvtcLDnX/l J/sdcnY6R8qIWrfS52WZdmdZ7zwu7bpztfxknypn1DNl3mQe7hwuPx7jJTZUaZ21unO7/L6fbe6Q nBselu/PlPNl2uV/M/iyjN2j0q47D8yvXT5jPKbe02cVEbXXS58fZtodImcDt8geI6z2Dqk5Y37t DpAz1sFWbCpTLplp9yqJNw2wzlPc+WV+uiqVd/taZ6vuvLOOjimNkHKks4amZqbZsdMRUo50Yqep 2Wpu2jP5JjPY/Hwr+9yQRE8ldKa/GhCf4WzODUMdfG5YKnsyPge/SGjylTdnbLiPZLL3kzb4uzLl 0uXiG5bKWa3hG7ZkN1l2ZxPnv1js/2LH/lMz8Nyx+75SJsdKMisvHd82i6+dqddevRq7rJQ17lNZ J//tssvpuH5d6jLZ5UQVo/a2v/buqTsqtvGfokCfTLGNsizONvj9DtpTP32u9tSVElN8W2J0X0r+ jq1/XnsPyjj0mpcmqDlJz0z2vjVfc1NxmnFn5v1dlm/qzti0aSuEtsKhTc3iTMe3wuJrZ3ZmkvPH 8i+Jwh6+e9g1DqfJWN8pex0up8u3mhxQm2aOlAsdGp0XatMskfIeh0bnito0K6Xc6NDo/FGb5hEp tzs0OqfUptkh5W6HRueZ2jT7pDzs0OjcU5vmqJQnHRqdj+o3p3BuXCb7eFf0+K6jx9RsVZu2TWjb HNrUDNZ0fNssvnZWaz78GrPnWmzZtzvn1aatENoKhzY1DzYd3wqLr50b69fnmWy4Vmy41rJhkzlr 08yRcqFDo7NpbZolUt7j0OgMW5tmpZQbHRqddWvTPCLldodGZ+LaNDuk3O3Q6Oxcm2aflIcdGp2x a9MclfKkQ6OzeNtrw7Via4stG3bn+Nq0bULb5tCm5v2m49tm8bVzgbOx1e/aWXmmNTLbmEWgKHlW 3v2KkjPWdQ7k9rrCXteLs13XnzHrOq/RjHN1Vl4vOWenJX/rW9caP1Xya/zW+Fuhq3nojWvgIRWK ziOu0n1G3iznAB/J+PuXq+0JuD4i+4rszsjrLE8jqfsCaT/kOr+3/Q/TR6Btss8bM9mwqce3NvVN 5mQ3DZDrblY9vrdpjMTnzP8i1C8N37DsQ3oaf1Ou/w/4xw49GEkAAE== ------=_NextPart_01C43995.4B4B8F90 Content-Location: file:///C:/0D741EA9/BIBOBCl(correctedt)_files/image002.gif Content-Transfer-Encoding: base64 Content-Type: image/gif R0lGODlhBQEDAXcAMSH+GlNvZnR3YXJlOiBNaWNyb3NvZnQgT2ZmaWNlACH5BAEAAAAALAAAAgAC Af0AhAAAAAAAAAsFAAAFCwUICw0LCwgICwgFBQgIDQ0IBQUABQsLDQUAAAAABQgFAAUIDQsIBQUF CA0ICAUFBQAFCAUICAUFCwsFBQsLCwUFAAsICwgICA0ICwgIBf///wECAwX/ICCOZGmeaKqWQhAM ayzPdG3feK7vfK8KBF7B4Csaj8ikcikqHBC8RJBJrVqv2FpBsSgwGotEgFBwNAyMwAMgAAMg5LQ6 S6/b770EDFCIQAZ6ABITXAkUcFxsaxIRC3iPkJGSAmsACRUWbEEJboxhFF2Jf5KkpaZLW45sYE5Q lCJwmpagiqe2t7g1gXwKUJ6pvKEIZUHAucfIyK+CjW+bMGWLjRcTUIaqydnaPgE5EnJrsWxjLS6V YmMOahABbtvv8DXdSq3I3y4utLlbUPFJ80k8JYvGx8EeGQkq5VGIokyDfv6OuEjC7sVAB+fcxVjW g2OKehElAgy5gyCbg4LS/xxs0UCOvhHlQHlpB8ULAXVBKlLA4IAAg51pHvJ5QtLIxKIl1blA2acL RkHFGGRgKAJISgK93jTgSZPTAg0Yy7QTprULUaQ9jqLFYbKNqopLrxakKiGRiGt8GDwgWNfpGoJ4 vew9uzYHvpGFZ5gUU0kciXs36TaDRYsYXy4ECUKo3BNk4huHP2t5mrLxngKZRqCjm/Wum2iX/c61 9DosYdE0QuOWkdmMqm/hfE1hR5VN5Q0MYPSN3ftvckGYb+9ecVjt9BP38AlV7WIKhwNLW7ws6ALM PTcs0QQAJV7poqCOWEK8nqI6Yvr46dvPzx+//fv9BYjUfwAKaOCBCCZ4i/91CjYYD4MORpgNhBJW iAuFFupQYIaGbcihPB5+CKKIGmJIog0mnlhfiirOwGKLJOADIw8vzijjjDvU2OKNOJbYIwo6/rii kCUESeQJRnKY5JFFhnjikkzG6KSIUEY5QpUK8mglDlgiqOWWoE1Z4ZdgoihmhGSWOSKMaaoZgwdw ugDnnHTWaeedeOap5551dgmAk36COaecfBZq6KGHnlmflE26qcOgASAq6aSUKgrklTwGWiaklHbq qZ6WIompjJpuGmekn6aqaqgmzEOgozxwquqsnXLz56uwPnoqrbxKaiuBrOYqgqy9FrtnWsAyqdQc RhBr7LN2IrvfZ9k9AMH/fEXwc9d4OCREgrPQhuuBtG1qmEWBjFW1HULFmeBQP3jt4BG44j5LrhLB hnmCTSR4O4NHH50VGAM3jcGLenE89A0YOqlCb73FIstEvrlhyAm2TQRVE8E4vZGPUz4BRdNQ8Mqk zkOcYADfxdr2BRZDD0NMK8Ui0EwdFbqdAIFGI2i7M1cog/EyeaxktbNZJTvCXFaWZbUcacPuKvOz Ntt8x1EmHm3CwHs99XQlgFWmF0iBeY0Z0505jRnUAMQ8dapVw5OphxebsJnSnZkt22x3F0SGwJwt cnZNYam9d9QeEPp2r3FroyWL/Kr2gFdzMUcaYLUNlvRsyxFMm2ws8QSz/9SL89o4Ml/WyM4UEmTi hXJrC753Zs8tR9SyD6SXhsgjexxABYmI5zDppc9atdVLpKkjOuv9Bt84Z+y+gPDlvDcW9Bi721oN bhc/6fEXbliqHdra0L33iIJ/Sorj1yHWujOcj76h6iOxJIvtvwMpsPz37///AAygAAf4v4rJIH/b 2B8BF8jABjrwgdNSQbkahSD5zY9PtzIX8prkoQlSsILEuyClMmiYP+ELQx7koIIsKMI8kZBLJvwH hFLYqg1GhIUtvFNaYmg/xFQJgfHAYQ7rtEMbYuqIHYqQEIc4pyKe8IVcAmIQQ8jEQtGsG0asWaCk OMXEoaqKhuJhCY1Iw/8YcLGLigMjBq8oRhqV6oxo/KIaj8VGq5XxgHCExxKZeEdRXXGGOcpiUfY4 RCzmqI0gKlC+8hjHOVoRkQZkFf6SyCFC5tCQJYSkBKcULEY20pF7wiQMNYkkTpJykyKyZAtrZq5T IvGAriwliVQpQlZmsoNJEqWLBFkYWl7QlqNUJJR0iccT+RJ9V8okIvtYpFgyyphUBCWddgjMW7Fx TdD0ojRdSKNkQvGQBmzRMb1HTWtKBJvZTOM26XTN/CnKk0gZZ/EoFhpeFnNG8izdItViT+r0UzT5 XJwkH4cHeKIloFPD2i5dFaPkobJHCJWZQr2pQoo6c5Sy/FFEIQaQjlr/Z4Yj4aUwibTRenVUi1hT ZDONYtFnajSa2zzpfyS40iKctIZHKqm4pBTBSzXUpjd1qZB0Gi6eVudmP7VVSw0KUJhKc1QzXVRS y9lGpn6GqNCy5lHN6EOWWtSqV3UqKM2JolYBNalgDas21zlNcE7VrR+MElaN1c2a1hWnW5prsehp 1on9sz96ZZyP7NpDNQXWdK0krFHSipvDziyxb12smxxrPMi29KyTFasaPyoPUd0VpY6iLNwUK9W+ BhKqodVsFRl0JpUGMqUXxZFoPXWf1vq0RFv9q4BmW6nauqi0HcqZbgPE2xGaFqmeDW5uU7tWtiYX uccN07SGC1jVMvG5/8CNrplyFlt8WneI2L1teCvG3e7CqLi+sixpF8rY6jZ3m+pdL1fbS9zvrlKZ 4k2ksE6A3kTFN7K52S9/7XvBTppRwEfoL/3wm18E70DBfFrkgR3sAwiHMrvyBbBSDUtg9ImPphP+ B3PVOccPY3i8IA6xqd5b4hTjFbqwjDGHWbxZF2dYww0+sVw7XDwY4/jGQL6slSzMTR3rqLW2zSyN wejjajY5xzpmEpHx9OQky/jKguLx4iiJ5ShTuG1afhuDn/xl7oV5amO2cZlvMGU7hcpS1B3ymTmK 0d+uWVdLLqS+OnvnHLR5UHW2c5/ZPGdxmfLHUEb0l//sxStL0p+Bzv9roaElppBGOq4BVjKJLylo Aw/p0jvOcy07nWYt/lfKk6ZrpiUMJFbPeNOjtrOrMV1WTcsRvAHGEoD8FOf8tLmsfEVxpl996/vm Gq7ntDWTa33qQatgym/+rLPjl+qZwZmq05YBkaMt7WyvYNvMdmuvXypqfYJGqfRVkYXhbEdTe/vZ 1YbbtePmzhEX23vsNm8pszhu+ig4l06OIr/tncNz6/vFySZ2wfl8cKEmfMWw3jI2/+hQZcd6l0IO 98OzXG6ZmUnRKt64pDter4+DnMzdljPJd0peEb9b2/GWlJGA2G+Sxjx96Dxtxl9egtkG6Y2c5fm3 b75gjHODtUIf+sr/95pzLp/83aJF8tETnXQwLx2xRvfyK7VrU4ILNOtaD/jTNT7yiGcV7FQXtP28 jma0p73LXbe4x93+9pB7VeFt5+qwm43sst/bXnSvO8r7rnKz9yrwOf55xfFuUkBdm9Zqjjvjw+XB +0E+8huG+N959U2EO13tVe850fsUoi0+OvQDvnpvIe3GUodep50fVSRqPtTRj8tEzMQX6lNv+Frh Pt0N73NEM577xe9+BMN/aGJPt3a504r1gw3+zlMe6t4jCvpRDPLg+V571R9L+XvWPuax7ffnC7OT JiZ85jlu/TXuWtfj37vIC7/59L3fnsyff/XrT7/3V4H2dkd/n9J5/4w0UD0leZrnKd8EfP60Vcdn de0XLR/VJQV4gKi3R1+1T55kgRfYYRklXePmgA8oRJ7XdOdSXmjBPCgxSN/lQwBILi94BGUgNlPg LkSgBTdoBxY0NwNSfMcgAC/RN1tTgzMgBXggP/zEboKka1zUa3UzAjuTAW4ABzOhBg6hHmvgFs5Q he1iBd1zIzN3RkyIf3RQBivYAlgRH9LwG4TwCYighsyADXXwMFrVRz5ogoi3fttFA8BBAsRwDcAw Cq7hCJ4AiKKwgnQgK1A1TCeofggIgi9SF/NxMVaBFxyxDLFQibQAMIl4Khy4bwBIgTjzWlFlAn2Y MWTQC8QQDE2QCP+pwA+raAxXA0G0WIu2eIu4WEBVQQvfoBzTUA2zoAqBwBe/aA3csj4v5z/7ohTb gQ4FYy3XUw7M4ozqAI08IwkiOG3JIoPS8Q4xmCHZeAQC8YDeaAUVgYg9kh0fQ44S0hvoiAL+ogtd mALxyI5XsBjXmAKcuBHziAL7aI+oQBoCcBD3sAcssQEL0DA9ozFeoDBf0AEfswUI05AIsDAJuY4A WQViERdN0AjRIAHFwAVD0zNGUzS+oAAG8BQzAQYsYzgjmZEaKZBTeBgD8A01aBLbojRjozai04qE k4oneTgwyQSLMQeOIQKQMRtQyBlACR09yQs/2TKxM5QxWQl9KIj/qMEdXcMQlAMbPHk5aLMXhoOT VEkPKukbpfEGFTkcW+mHtTOVLDEVqPg5BBE6bFOW4igH5QERzggA38GRwjMC5/EWLgA809M8Vbgd FWGY43CMeFkU5fOYBhKZktkf75M9lZmZfTaDcqiZnvmZoBmaojmapFmapkkSG3kYe6EXMiCJp3kL QgiSNCAG8POapCCEN0CZtikJfXMBSuM0eomIurmbkAAXrGAGw8CaA7kv20OckdCbv1mRiRAvftic zvkIuEkymoGOw3mddpCd9XAPjtmd3kkH4PkETaEC5FmeVrCR62KXqlkC1VIQjsme5JODrWOf72AV VdGP+nkKVcgsA9MWAgA7 ------=_NextPart_01C43995.4B4B8F90 Content-Location: file:///C:/0D741EA9/BIBOBCl(correctedt)_files/image003.emz Content-Transfer-Encoding: base64 Content-Type: image/x-emz H4sIAAAAAAACC81bf2xXVxU/3x8t5dcGX1iCA5MvWpAgbQAdEkMEyq82q4BdYWz+AEa/Yhf6bUf7 7XdlE8tA36PbEIcmQuIXKUbL6HAbJihrshp01OF0LhEwWYgx/sDEP0iATKcJnnvvOe+d9/re90db HV/y4d5373nn3nvOeefd97kQAYBdYH7liCuIm+D+uu4HuD4LILnms2sBItC/GKAiAlAGvl8cEQNY gn1/9XXdOB+HtmdigApgxlSASYh1q+tWoWi3uuf0u+cblNytKWqE8ahnT4WaTb2+W400MRqBhK5N ir4Dg7oWj0ym3vy18VH8K4oK4f1oGep3p1uFF1VY1kRwEjAPUUGIkFwSgn937tyBQj8zQkTPoTwS pdlEdU2NVRlL4OhzYb5AFcxCfAiqsVyE1woLCaqvmvBx55452D4Hpmh95SQ3V7cxFiLCbFMeUeuc pO+OO7NRIylNhW1r1sX3V8PHNKoIbvtpjSqClDeypr8Ur/prbNFXUc+r0OdgAK9fRwxAP17/mHBG Y4DaTd8pLX9W47RjUSN7Wre5OFOkRXk2A7o8U7JFqxyr9Wtw+yvYZtCv4ZXvp/ZCFsXHGv4dvUWh fDiiKrFoJNHY3JJqT65PZZMNrS3b03Dulxezg4jk4JtZwMgwkR2leFbrVdZaPOEV6NU5ZEOmoy3T UYH1OKIcH3F3zHu1ZDnM1G11aRTsMFJeubBa1GefWkKdBrcvwjaDOg2vfB211waOYjKF/zmOidUq iTmxKCxFDSsQfL0O6xvE9Wasf0lcp7DeIq47sP50yCw4Pw1f8368w6BOg9stbDOo0/DK11F77Qie LdZk4922M0qtaK/XsAhS3sia/vB1BlnXggbEo+J6G+Ir4roV0Smuv4o4IK6fRRwR10cRJ0q0tg0/ 0rAIbvtpDYsg5Y2s6R+5tasgQ+jUcGM7Q+jU8Mp3UnumRGsvxTtWINxYzmAsZ0QsZzCWMyKWMxjL GRHLGYzlTMmxnCF0arixnCF0anjlO6k9MyrrWgRbw7WuRbA1vPI2tVslW9dC61rCuhZa1xLWtdC6 lrCuhda1hHUttK5VsnUtgq3hWpf/2BpeeZvarVG9j0v1x0lYrtFLqNftlbrtBGxycAyaEY9rfBNt wtgHzzloxd404fPwA8JJjTRmAUY3ZgjGIXjKwVEcg/F9HJOh5sX7AqVzDbxIOKWh9Lvx0k/w7nlY nt/U/w8b57P7d2ClxrcJbHfVdgTjkvE82puxH+29H7OtQhZbGDvR3ozNaJ9NGic1vozWZHSijxjP oC7Gc6ib8QJqYKh5sd2Vzhq0ocEpjU1F2J3l7xa72xhLChaB7a7avoErYuzDGGR0YWx2oZ0U0mgl xja0N+NzaJ+NGic1tqK9GS0Y54wnURejG3Uzvo5jMiwR70rncrShwSmNjUXYneXH1u5m5zoFxukd 2oDerUWi0Y881NXekWpRVxNgNvRE6Fvtzss3E/Czn1/MmlwZ1znzBuZK/w4YRrwDVjOZXRaBx9CG lWKm3L4D7RzUvh2Wedon6P3xfNgLalErkxUhu2K+vwa/2A4HtK/C9t7AO2frEWbg2xM/w6Eh9USm eXeqKbmjNd2xu3VXO75zymg37u7e3Wtlu+Gj1WFOS+SdZxp3c4mA9jaMzKB2tbsIau+APYHt7Rit iQJRY75SLuCOVH2lrG3emVxcbb5RYqFfHzP0PWfxjTxRMQWp9o5kq/68SSs7leFdMbJOubaOms9/ ol6WoYziLQL++ItpqamUF9TbfiLN84+I9xR74bueQNcJzVK4/YOIdxDTSdfkIuszqK4CeY47j+4v Ei+kxlCRWEnf2PUuK5IshTHJ9+NxlYZpVI/SGrm+GdFET/4sms8CkhkrFkWN8Ulaw3zER2kMFS1h zEqYXaeJNW0mPbNIN+uNi7kz51KMz6b57LKA9FbR011Fuv18jJQdItkhR9bL0QTpHRJ6JW9TasyF 1dnHryF+gXgT8Rb5eKx4Hfbxa7SeAcSvycdhXE8pPua5v0F642LuzAKNxsdDVP7KiR8vQ8Sy50n2 vJD1s0Z+veeH6XWZpGLmrEjctoiZ8y0fRarerQ6/hO/IFsyo7TjyeiyzWDbgrr0F34bpEvJG9cMX s1HEP8ZF7uE2zkfqN+VKQpeXVv8r+/b2RfG3nr2dfXLcdPjU4dt6J6B+V7Fd3b86esgkY1zR0kdn QNPlBPThfTB4KFax4hPxKW9/KwbdWO9WctPh0z23s0r6kXH3dQ1+4Q/ZtrJL2o5qbFeX+ck8yby5 kv1T21C2jPpv4HiRbrT9D3EcHVHTofGymX8truOG/X52Evb1XToSazKkh5M3Za5UdzYqPyD6EL9F /AZxjfpeXDu55upfJtX06ngwcV8/nN3Wvy2Ky8M3cQdqU39XkHycynJD+XtioJHeGxOpX63tPt/Y M6mvrMDYdRgNZmTvmEHjhtUTIc9RI5WbnHj3MoksW0mylULWzy769VYO0+syjmHznEvvh1ie92PM 5+eHxXvwIcQSKtVamZtkmc9Q3zohw3wly2ygvkeEDHOYLLOd+h4XMsxrssxu6ntayDDXGbZ+ufcI 8tkBsu0BYVs/E8qyB0n2oJD1s6N+vQeH6XUZ09G+0+SceqhU4/j5VCm7hWS3OLJejjVI7xahV/Ku hWweFldK91Yqm8kfzMpKmTYquxwZw9RKmX1U9jgyhr2VMi9QecyRMYyulOmlst+RMSyvlDlL5YAj Y5jfkcae0nWB5n7BsbGXF5ayQyQ75Mh6ueIgvUNCr+SPxyL2klqvKXucvORll2W+s6nscfKdl3H2 660cptdloUcaewtI9xIqTU4zHLXMaTblNNvJaYa3ljnNppxmO341XLbMaTblNNvJaYbfljnNppxm OznNcN6jyWk2lT1O7vEy4jJWbPEMmZzmZcn9eg8O0+sy52MVVzkqjzv+9/LqMq5yVB534srLtfv1 Vg7T6/Lvo4mrHMVVzokrw87LuMpRXOWcuDKMvYyrHMVVzvGrYfFlXOUornJOXBlmX8ZVjuIq58SV YftHE1c5Ko87/veeBci4ylF53Ikr7/mAX+/BYXrdM4Ox+hYcaUz26X0j4FefKdUc/WcLm+n9sUnI DpI/7rbzBv5eXYZ4neZ8kfbAYWcQ+b8XvTytlH2DxmFZ/1nFB+XbQj7/LuJBxFEq1dz95xrs8/VC ln1+t511sM+Xks/XC5+HnX+MxudLhaz/nORu9flBenZ7xHPuP1Nhn9cK2UFnj3N3nbOwzx8gn9cK n4edvYzG5w8IWf8Zzf/Cb8wNxYmvHie+bwfoX9voExy0eju+AVPQAiP9FcOzFMsNgYfP+WC5oRvn wrmhf54L5obipIP3CIU4OhhDjm6s1yT9ECedXj8YPdd/n9Dlu5eNHnvW7azilJRvL237W9ZEJsA1 rK8fP2HD4HuzQZ3s+f3wU7xfzbn+agLG1/49ay++rvWo+z780sWsuk/NT5X6nBTnfa1jSJ8SFsPR NRI3tgcxk3Ka5Mkq6bs7kocn24j1xyhP+Hk41n+Y9B8P0P9UEfp3FNC/BnE/Xfv1LytC/3ayuV+/ 4jP26rMXk7uX+/TvpXOaaAH9KzFSKgpwhryeIzTesYDxDtN4hdZTQ/ktSH8f6X85QH9vkfpXUZzm W8tP9HeE8fsVoZfHytDZ39Q8Y7XpfJGCJ/B7sRm/VFLQhJbcgVkgjbp3Y7kL84P7zEhu1s8Rh/XH CUG2Av3uM/1bffNPUNwVslVdgP6wsbbSnNIBYzUUMVY6z3Oifq3UvzdA/9Yi9LcV0P816j8UoL+l CP2ZAvoPke++F6B/TxH6O0LOC1h/juLkpQD9VhH62wPySD6/t9Ez+Gey/03fuLXEzxU6I1mLT8dO fDIWQ7XnfCRWxFlFI+VnlQN/hzhBZ/1yHiniNCfmmUeTzjspvXdKYqS5pzbu81lGc4n5nsVy8Szy +wplu+eJ91XYfoLbMd91z3H/LUH3QqrfK9pxvd0P0v6P/2fHPMHXzBNj3AP8ljbyqv5fUqZOnKwy AAA= ------=_NextPart_01C43995.4B4B8F90 Content-Location: file:///C:/0D741EA9/BIBOBCl(correctedt)_files/image004.gif Content-Transfer-Encoding: base64 Content-Type: image/gif R0lGODlh5wAVAXcAMSH+GlNvZnR3YXJlOiBNaWNyb3NvZnQgT2ZmaWNlACH5BAEAAAAALAAACwDe AAYBhAAAAAAAAAsIBQUIDQ0IBQAFCAsLDQAFCwsICA0LCwgFAA0ICAUABQgIDQUAAAUICwsFAAAA BQAFBQgFBQUFAAUFCAUFBQsFBQUFCwgICwsLCwgIBQgICAUICAsICwECAwX/ICCO5CgE6FCubOu+ cCzPdG3feH4ThaH/wKBwSCwCBYcgwmdsOp/Q6CihUP2Q0qx2y3UtGI2EI/AAQAIFRWBwijQIgSSA IBmvu/i8nijoARJgEAMEBwkTFGVnBwJlZgMJFUx7k5SVLlSNCxUCKoJfDSJIhqBfBgSNlqmqk5+h FmWABowimJo+EEkXkqu8vVI8PlQYPqIToHM9EGUEFo8YvtDRRpgitn8TGRaRAAJugA2aExoODwRu 0ujpNXAoEbKNYhEbKGjBYwXj8nHq/P06o/4CCiRibaDBgzXO7EPIsKHDhxAjSpxIsaLFixgzatzI saPHjyBDihxJsqSMABqp/9BDIeefAlQyWnFZ4MAKF5Qbu/mA0xIHtRk0z22heacLTo3mmEBw10Um EAI2X9CMmgUFx6Qilu7K4tQl1RZivkahtxHrCSsqyfyxE4HDmAFEVUz94rZHWnhsj01RU0/lOZoP 1BSyUwABg7oG/IIKO0IhUydkkdLz86cKAMAv53irebkm0Q73glmeSq2b3splpirjFgHBGDefwhKl vFon48vPACx5EjljUp3V7KCwwDQ2Z8aywQRfWa648hGloCsX8wA5GMYyW1G/TbNnk5VHLyalIid6 KD/GVSD3/Nz8eR9dgY0wt/jl+kecZdJHfbvzwu8rIVVcfpyJgFV6a6nH3v8xU5EAXFdONUgFfgpS eFl7x1XRX3AwEQGeRsBxQwYVTDFDzoVhnHhWgjKRuNMD3aFYAgR+JJCBAj2Uch92z1GRIwOJFbjA MbMA2FtFad3h140szUHPOScEAFoEGvDVzjFpycFOHULRYiV+AXijRgQZvMZkHWFi+VoYY5Jih3dD gGeVSSR0RecLcoZHp513tpBnn/49BigLec456KGIJuqhoozi2eijK+gJaaOSTqpopZYiimmmg27K aZ+efkpnqB/xyUKSV9rgmAG6OESqRyd0+AJwJJ72ghg1OiAoQq9ylAAGNG7lAladydqCVlMY42qf UO0nA7HUxXSig7Ya1Ov/RrpE++xjpj1b1ETXZvRrVpTBwE47wrawIkXhYjSkgV2+QGys3oq1rEmo qhUDsWbs2gJgFbVr0QIwBbuvoEtVu4KLJEB1L0mG7AIwDCHyNEM8TBDykMATiZHmcvWAZSU9cN5q ZcnWisooxypzxHLLGr0MM0Yyz2xRzTazmzOgOO8cUc8+bxz0qEObBHTRvCJN0tE3Q3YR0wE73fTK Ukd9aa/hQp1HlPqOoLVEhsKQNUOMhSjC1z/TMLZDZ+yKNkTXro3QWfwC8LbQM8htUHR13/1wDHH7 /cs5yHpNdd5qCx5FWhZ8fPbhJ6mdikJdC0FqoZti7oLmfha6ueeAS454/yqm/vDhqKeLHbcqpeug OK+vhi02688tevVJq5Nee5xXx17D6zG0ngPwAx25+e+0K2z6paKPbonwOBAvUO7OVwL9DdIHRH3k uivvOvPVqz45PcZ+n+j24j8NPvfsT61p86HTvH767Vutaezo63y+7/Db36nsfkKe/O7Hv/Dp71Ao KWD9Dtgpuy3weBnJXj8S+MAAso4cgrEdAh04u/5RYk3XMyDqOOgoEU7COt6zgQTVgZPLeZAS9+Hd BnH3whMep0BBWGE6sJY4XsTQcgjkoQknMaYyOQ4IOpRG/iAYsxkCDmtJNIoTO1hB/42wipHqSBR9 sURCbfEmDTScBbHIQP+iiZFQNXTfncKTubx9EYygYmPnhqhGo5UAU11UH6juOEb6uYxnfEQjGWFI jjGUL3pxDOQdhQiN15giXstbYxYn+URpICczQJSkIknASGhc8pAq3OMmz0hFT2bIXtgTZSlX2Yto 1S2Sdhzk2d4IFgc0DpKwLEkn8UTLU+FQg0tTIC/TsSEZ6pKOXutlC+yAykv8kpNmZOUcLeIxeylT D7uc5kWKSUqRCPNz4npmN0ECQEH6kZqFrBwJvVlCKKZEV488zTXx4KkW0vCd6sHk40YSKgreM5xW WM04PdLPdYJzI5eMyjy3UE5O4k8gb4kLDRjTN35W8qL+wI44Z4WuURL/NH7n5McPc/hE0EWKcyc1 6SL/1DmWSvOkAxkpEgOkxc8Zz4sGkWkutehCjAYkWusiKTlBOsyDRGlK6UrlUF+6yDB+lKkOneIf oTrL9xG0oD5V5VRLWFSpRhCrIdWky+p5uYWmwqwA8qOkGirWr5J1lGwdiMOMgNYixNV4bIzrTy1D 143oFYB5jQhhkjo8v76VknoVCDcvxpdyGTRgh51kYhW70RhQo0EDBVs/8Wi3uvqksjBwj0fBFdmU evYGi92XY6H5tNKmFEmglcruWAvZsCaTmmpYrQwMZiPassu1TUVUkhR6M+BG9VOnjZ5xZ5ncXpTV paZNXXRp6kXoGgql/zgVlXQZcley3vRRze3hQbOrsvD+k7wwhZl5bVvVls5svV19rTZbBt/5phe9 +8xUfe+L32RWar+77agxucvWhn73sRFZQG52M+CDHLiz7u3jRGLUVwJnDo93ddcYUFZY2GEYw0Qd 2LQyaS0Dm1iWB3kXN0AZSgd/uLrIZAhR/kPi6b1YvigGZPFuvFIB6nfH9vWv0nzLjxP3V2nhNXKP hzzaaJTzuw9GWnJvCuXJBu20VH6yla/MQsBqGcBJQ0eW/wtm7orZyzBuMKTMilctV/jH0GjzWr9W 5jxcU878fbOllHkk6jJ3LMjlRZ/RLIU601MVg5bdlpEYaEskOrgMbf/0JBKdV+kZWoqTDlubsSlp O19X0Zf2KzY1TWom88bTR3m0qaumBZpSN8qrZnSryXI6P8f6O4WmdW9sfWtW03VOkdlur6GgtWDr etHDrvFMrRIgXie71ZZjtkqnxymZYQ7WE0Q0lrF3bWTvUNvYpoR3u+3sv4kbuqtIILnDDS5wT7sS 6463vOdN73rb+974/lC8n23hd/N7x3L6N9zKLXAXF/zgCE+4whfO8IY7/OEOGS7E/QGHRsChmRO/ YFRw0QIG18DjicoXLtUhBu/0QVhYsEHKH0UrBYwcHd0igUK0hKY1qIQNabLYHAaTAkgRa2Lmwviw hN4wVJKnBPEokiD/IMEEQyDCDHFQunq24fPHaMtkRF8YX2UwIWd2qOLKKsUppEOkQoQdSHNgMaDm dY67QD1M40BBuRQTl7h4LAB1eIAh6/6Wu+u2WCTwEQK2wfFWhQIeL7EGxwFgeJ+vhCldp0kHFhwk qtTGHRpVkN7DFM/MJ+gSCuiJaVbDDGck6xjAIH0z/pCbSc2LDDMejgNasiHtkCOGYfkk7mMb84R9 IxzjaA44tmGa31dAHBg4hbMaxa+lGEZiG/78fNpunxtWSPfWl74LuCYHj+FDV/OQA9dyhYZ8yILG zEdYW8QJmNqfcvcWmhD8p6GsolUsDgxjBpGq80sfGUBHKqIkNeFK5pgXgBViBAXhM/myEFkyY0mg EnO3JqGAAkilEnVwS0RCgfAEgYSlKk2ScTaQWiAYDSI4gq3ETCaYgiq4gizYgi74gjAYgxGxgIGh dgbygTI4BCdXDTdigwoGeDkoBDt4Ax6gFH8XhCrnByBXA0OIhDqwg7qAWf4hd+myck6YA1zTGm9B CwMYemBRf1eIhX6QLUKiHE3YMDYYhtunhJXnJYlQMkynhk8YgTZBFP5yDfCRQnKoLnRIC1THApTz cnt4KnxBfR+DKhu3EnaBIx04iFyXAdDReo4YBYvHDVk3iT9wd+p0ayEAADs= ------=_NextPart_01C43995.4B4B8F90 Content-Location: file:///C:/0D741EA9/BIBOBCl(correctedt)_files/image005.wmz Content-Transfer-Encoding: base64 Content-Type: image/x-wmz H4sIAAAAAAACC51WXWgcVRS+M3Nn7s/slpWEKKGUDbQakASk5EmKWlitVSKtBhSk2MZl27TJFLth cSlEigYJiD8oiAEfYq19iNLS4g+oJVAtLCyNFCsIS2Ih5MViX4SilvWce/fsna61dt3lcOZ89zvf uffOmbnzU/37Dxjj8O/NrWVfYPh7ZafHFGPB3RGODBksANsEEWO+57EQfOxjFl5l/F+yuRyOCS9g AnwfX8v25rjRZSyCDJvpt2LfqG73UHUQTLbMgz/+8uzWv2azyZrob2NYYQhkcGZQCMrilfIh18ca f0BsqzDfcj2mwW8OAnZDPOxfBUO/2WC5FnYx+EY8Ge4R70Ur0ddiJaqLXdE5eSacUR+Fk2p79Jlc BnwUxj8FXgP4pJMzOhFD7u/8uloztm7yEKPag7xfD3Kl0W8wWMiISzjGxD+ry/5xMPT3Guwug72r fT6ik/Ck+jI6qWrRz3JF3BA/yN+iGbVibFJdFedkTdbF+8DZAtzXIIe0qDZyz4TXwdZNDsZUuwJz qcCcKqm5EpdwjIk/pt/gBTD0DrsE2CWOnjSIRzjG1HP23kXUCq17ZzvqWmt8S+Czgn4wLOizoYuH ooL+JHLxBlHQrwsXX4O4Kl18RRb084o69t96xtb1mVvPeVUAQ+8wpQtg6N0aLY/wNL8hRuCujEj0 DqsCVpXoSYN4hGPc7T7VxRXI736dDXFe1cHQO0zpOhh6N0fLIzzNL+ijcI8nOHrCSvwoL/EJjp40 iEd4ml8XPbDuHoneYSuiAYaeNIhHOMbd71UR8n6NXPwqvAF0Kl4I6+KesNu9rItTvAGG3mETgE1w 9G4Nlkc4xpvMWJZV5XwwJy8GDbEINh9U5RhYqR3PyW/NeBXsoMTx+YB0kYfxQWMlk0dvbMdx+sSn tzjN+XS8NTgdDwToCZuTi5CHuYvtesQjHOO8GdsAz8VCOK2/Cy+rjdEX6qHogtJRn94T4vsJMYyL 8Fwjp6gvm2ccc0gbeRgj3qdnTF7nWtI1iN+5ljH9NugvhOgJOxY/x9+JJzh6p2V5hGNMfYV+uHVO 9vr7M3XRC1d/+g80d5mGeMs0BvdZ7pkDk8Uj+dFiJb87mdw7xagD7ZmFcnASs4EwYo8eKOW3DqMW dZk712gPl2Oll+MfVZJR+kOwJNMP8SjYNp2043WFnFVjyHfPLPIwXjW2zeR17mG6BvE795DOAPSE LQFvKe7XS6l6xCMcY1pLhdf8HWAVngG74M2C/xziEy0M42dbnB38K8DQau3zy/JqBj9hrOZ3riVd g/i3WsssVzetBXmzLX56LbPGqK7dF7/dByNyX0x9sLvVB95t+yBovzH+2Q9PTZcPT5dR06of9970 OrvM+48us+oBs5ohe3wKJBkqWc37dUaS5s476lw3Y2k0OXssSV5EHav4sT4VkeLoHT4LNMusUcyy 4lQyXdqfT8wGMFS02gNBT3t/n+hytrHR1uyR8fHi4fLefYeKqGZ1V4Nj8f+dsza6go0nU+WXkkMM tazqfDDsd/dWcLO9+X6hkj1x3Jdy50mGX+V/+fYLmkbpPGIsZ+axMf2pbeby9MtHysVJ901Nq0J/ n9dsBob6N+Rm4JtEDAAA ------=_NextPart_01C43995.4B4B8F90 Content-Location: file:///C:/0D741EA9/BIBOBCl(correctedt)_files/image006.gif Content-Transfer-Encoding: base64 Content-Type: image/gif R0lGODlhUAESAXcAMSH+GlNvZnR3YXJlOiBNaWNyb3NvZnQgT2ZmaWNlACH5BAEAAAAALB8AJAAN AeMAhAAAAAAAAAsFAAUICwsIBQAABQgIDQAFCw0ICA0LCwUAAAUIDQUABQgFAAsLDQ0IBQAFCAsL CwUFAAsFBQgFBQAFBQsIDQUFBQUFCAgICAUFCwsICAgIBQUICAgICw0ICwX/ICAKQTmIaKqubOu+ wSvPdG3feK7vfC8SBYNv2IoRj8ikcsmkCQ5IxOlobFqv2Kw2oVgcEQwhVUsum881sAGhCCwSjcCA 7SYVHIBHAAognEglXkRVaIWGh00EEHgJDBFhfgCNEm8NAw8HkQATYgJTVISIoqOkNHBTmBMOCRRC DxUWIyeNQmCMYUklpbu8vQBqIxdeCBh4AoK0eYsATz/LSLq+0tNniowKU5GNeH18nnnCkpa5JaHU RefpOFxTxLcZc8V9d5JhxBQRGraqoNHq6P8CuiBQDoomNosIluAjqQ2EDQru6HEDrZxAFuYuapTB SkyhchkFhtxIUoU7QyD9/20ciUhlyRuAnp1JyTJdTZQ3X0qj+TInGpc6//Es6XNm0aCkaAINeLTM UqQ7lTbdNVXLU6i9pFZNGhXrxa28wF656tWX2FJnm5AtGxZpWiZr2aJ121WuTbpm7ap725KvEr96 1QaN+zHwOcAfEY+pMdTw2MGKBxXRatHxY52EjYqgnNkyNMhUOat4EMQQG0EouKBeCTox5RkkPpmh s5oLRaKtV3Sm8VrF2kkylcBZneK0yS49c28GyaO3jLV+SHsUjpyF8eLVSUY240+pDanNObZilwLO QhHm96CPI+cXtjhBbAe4o1A9mwHwayFPL/suZtEYeffZC5iMQI84J+hRSf+CbnTk3gUllCbAHcYZ Jx0bElLYhYPX7fWfaOBd5sIxIihY4oEnclNiSnNUd1qHfcy3RnWqnTZROf1Rsx0ZnFU20wt0gMTH A8ERqaIyR17HznUEzWEPjdjYGBxTrWlV2ECyTYiHdClwiQJpSUK5gIX0qHGdmV2Aydp/Ae42oHUM hDlLA3xIAUedA8CxzAMterMImUIAMaMXd7pXyZ45tgBIez+xmZd1bZTmXgmLyLeMpdecxwaEAVwK XwTsXdDGAvXNEqkB8jFEg5eN9rRjD69CSpxpuPQwXA6xYuEmj2fAeAgwtmb3naNZzdbGbYgAw04c UNyX30QLTDTAsSjyRiz/VZ65oIZtQZDGQYRCaFlhF75a6+qj06hmBTAVPinIi8jBO+sMuY5VLw73 6mDbvF/gMu6YULY4DLnCMnbtXGbcSkO5UfgbL8HIQGwow8+dW+wZ6i5ccBJc/suGn6tgEyPA/MJg cVs7pMdHepe+t4d8BUA07R0s40HxEIt2EGkGOyvA6TOAiFoAqPMdSe/J2Oo7joKniANBBKdemGaE jIyjpwM3Y5G1D/kusauuPKj55YHy/uKuoSmKUHYhWzfX49twxy333HTPjcJW5RwJBDdLPkwy2vPw HeXGWgRZMtdIc4UruF3eVqPf/64Y8d92df2X5UfjW0Khv+TZAD3WeOz3/3qg/0k4VJiDgu1UlWEq DqUhz8fzfB6wVw4y7FV6auVEpY5OUz5mC7Z2vmMEwFHBCy+iRl+rhXzzyuPae2jfQR89vtMnTa/1 199QPNff68Yb990Pu1L4vo2PvmHrL44yDOSXbz7z7W/2XPzyG3y+9pPVzz7x7WNJ8vI3PPoF0AUD JODyRIK/ISwlgQq0gv8YE4P1VVA3DYygub5ywVF08G4Z1GDmOHg8rhACgiIUDP1K2Bf7LWeCwoPh /VyIEyPYMIQpHCFTbNgXXVQQhzmsGAlZ6BoBBdEpzLvbIYx4RO4MsYeNaSKvGOhBK0lxilQSBYCu mAUZIrCKTORiAfcSvv8QiRGLHtJiSs74oywukTlsbCMZoRjHK80xBV37YR1blUY8OtGLIgRk+gbp lJAIUnmHXI74nAiQPcJFKMYr5Bcd+cg++lGSjaRkLiC5SKsIUJNes+QlPWkyUG7SP53UVU0Sma1E lnGVpjyljtSow1gi7jC0tCUa0YUTXSJxli30pVWA+UZWBpKYRRTmMHeiRmMek5lChAsQlRnJi0Xz cs5MoQxzEhkUZpKa2LPmNSsCPHDuAIY+UYw3v2lO/fFvnA78Wjb/975agg9v7ZTeO+EJK7DMMzD+ Qx45w5NP7+2TnzpYJ0ILys4l9hJWDJ2fFh8a0WV6MDEV/eVEfzLNjKb/0qEc9egu7cidf4oRfZjr qEg/GtKSrlSjFA2n5l4K05YmdHHqpGk1MXpOzXVTpyzFZE+rtx2T8o6nntyRUeXiO5XiUaVLZUvq nArCWEW1LFNNy00U6kCgElKoEIVfva7qlZQupn8j1anl8DcSrp7Vq0Qk5VufmkckLOpwTcyjVanK BD2gAllszJdT3doLE6FAAFMySaJmIIWjWtR5JuWCqphRLZPU6gbAkiqPXkXYXghKBVrSAx8ikbNR 2WFL6unDH27HVLlCtiRXK48CDoAJTXBCFuhhACVOUdsp3JYZi8XLGL3G1ybcZzQUSYYt6iGGV8QC uMw12y2mg9UuKrW4/8YVWWoasKdlNKMPQEMGLozEjNEmFit7hexVJZuCz0L3AeFoWj24QQBvJCi+ 49ALZ4vaWQn6gCCCmJAYaHGPfOwDa/LYG3MLrI848aO1zpMmdu2Zg/qoyjYPiYgDoBWj86DHIRCR SCD0G2FsfvCihnBQBLupzsrUr3gnWXElyXnCg2ohJkYrX4tpHFSUwFWJ5OBxj0nqVcCwFYUv/jEN 4zmIdSZZyXyJHxyV/JUmx3PCVH6T2+5JVriexXpYzvKMcXrOMIt5zERNaH/PjEz1sc7MbPZvmr23 5jij8n7Pg7Odh9umokwZzS2hsgDdukY+g/QvjjTkn/uH3SebOI5VMP9jVSX92I2auM6IrFsYaxpo 4lI6f5oOtahHTepSk/p4cttzIHuk6iA6p9U53DSstYnpWdtamPWJFl5l0LZbXwQOByLBrl3AYV9j xb0lCq7Ghm1sdbA3Cb1udjoMy4KarQd2hwUJs6VNDQCjIEgHaNrVxM1dY3gXsNzeiF/LQ6c8kC1N 7x5TraKdbmkgREWFUnBDBqBvduib3vX2hbC3CwXDqsbg8BYDwAPOi32tBwrAxoM1Ig7ekJXpdAz/ R30CUIH95I5vH/9wSrad8XSNquQKXPgKFFYDlqNcFCpXQcZajvGXy5w9psJPamsGLQ7faD4yY9bK P95zE1Ar6DIquhz/qJVjm5enFe6ZVtLjY7Vy/0tdoYuIAViFIKdhDXJTkxFlvz4wyjmd2Cxq19bj 7TF/xUntXWI72MuuNomt7ewtIO+3/dbvwZW9aUaKHAr6LrCJjW7foiN5xsVW97Ij/G8WohraYPT4 xAPucWWPeb1jy6erf07iujsBgLeBHbpvt3QhE70bxgUy8gBY85s/Vrh7JiNrM6MEQoMaSHQmu91t F9u3D0Du4QORn3EjaBEhWmXxzoQEeODbcVIC7JnPBGv8QNk7mD71hWM77OvAcNsPv/jHT/7ym//8 6E+/+tfP/va7//3wj7/850//+tv//vjPv/73z//+ayQ9IBEfNSdyk+jmf0jAKgSAKgNIevpmgBwj KTjwAXxzWQ54BFzifNmnARWoBBeogQ3hBT+3fDG2gRYoJA3RIMhhfSqgEMtHgjvQgY3HLtHHAk3i gkRwgc9nKIB3Xl1ngz7AdWgzES14WIrngysAhKdBerwmD0aoNAWRGpHiAcdiAo0DfLHVhEeAgWoz g1hYBiqoCV1YBgDofVwUAgA7 ------=_NextPart_01C43995.4B4B8F90 Content-Location: file:///C:/0D741EA9/BIBOBCl(correctedt)_files/image007.emz Content-Transfer-Encoding: base64 Content-Type: image/x-emz H4sIAAAAAAACC81ba2xUxxWe3b1ebAOBGtOa2EkcMNQJfgGOF1OqXds8W964AUwgIWDAwd5d7F0W UrBNQsC8XCL3qVa0dR+iJZGaX00TKsWqLEhpSaS0NFIqfqA+0h+VkBp+NKlCzzlz5t6513ftXfsm yqLPM3fuueece87MmblnBp8Qol3I32TAbYb6PVQsxGVA6fK1K4TwiRNLhFjuEyJHOH4GICBEHdz7 h+PWndcMEX8uIICBeGKmEBsAK5etbgLSXnzmo6kD7yDd4pkoIQ/4/CBfiKBYQ0+jpMn+j+/lUW2K /00xTDXDN5Xvjl7L88Mffy78+dCfA/wtdSvhohLKRh8oIcoBuQwf05UK99+9e/fEWD8pwUc6BH1+ 1sZPNZRVFigQdaJQLNJQJ/IBk0QIyiVwjahn4L0Qo9Z8Zga0zxDTiV+Q6QqpTaEekM42QR++5xR6 2jC1QUnIaWzbyvdSz4fETEIdw2o/Rahj6PSSVt7PxqvOmrLo68DndXHCxDBcXwMMiz64Pss4Qxjm dnnvBaK/QjhlWlTSnqI2C2cytKjSZpjKM1lbtM60Wh9Btb8GbRJ9BDt9H7d7Y9FnRBiw1EQCrlOA hIjAdROjkZDgdnnvy0R/gBA2LSppw9RmoTFDiyptElQ2ZvVeTtuGGRGCal8CbRIRgp0+wu1hVyky yjhjgJ9Hvo8p5gb8wCMM8TBsXq+B+mbtehvUn0ojRcUuK54oHffCExIRgmpvgzaJCMFOH+H28IT6 y8iekxRHxWENSXEScBp8d1IchGtEJwPvJRgp85lj0H7M7DmS7ii1KXQCMus5SU0L5DT+nnMILIVI Mqz2JkKSodNLWnk/vU/dekpSfAWwSbveBtilXe8HHMyypxwCWyCSDKv964QkQ6eXtPK+Nz0l05rS ogdWB4huxhpqL+O22SaOwAhV6BLrTTwjngbsJjwJvUBhMzy1iXCMsBPsqdAGzyh0Ag+Fw8Bb4RjI VEC9VI9FnsthZpE4QUD+VkzpY9hnRkWv4rnX9lQjtFvMA19a6BYlgFlgzxLxHFwjjjPwXg/jqPnM XGifa76vpJtHbQrHAZmN0G5NC+SU7bjsEV8kdDOs9kFCN0Onl7TyvhcWbYEl7RYNLTBaWqDvbYdy J1wjdjDw3nbGVvOZHmjvMS0q6XqpTWEHIDOLtmhaIKdsLbod/IBoYVjtg4QWhk4vaeX90eWVAD7y f8BL6Qs+rAT8voLmto7WrtJ1ranSTbGOXVHx6+GrqTcApW/8PiVEDUe1gBYr0VoL88+Ly/DVIMT6 ZCKeTORC3QAERUCTOY0op4kCalsdBcKEpLLTSd2mw1ocf1doTvf5/bM3H+lKtHbgVb54WJzx8dfA vV/9p0D8ZuhqSupmkI53QEfnO4os33EW6ytjQ2NrV6J0dyya6Iy154K2OaAzah2kdzUc75B9bSpJ OwH+Q/krY7E9uWwbi0ba+kWgQVu3RmPJfftzmMpOOXbN/jXUIIoBD5oIwXUtICTuh+svMIoIIW6X 9x4g+gih2Bw7kraY2iwUZTh2lDYhKouyHjuN8JxECcGKcsWMEoKdvoTbiz2JRg0wXzfADKUQguta QAjWWQ0wt0l0EkLcLu+liD5CSGoW7SSeERs6M7Zo0tSiYZSn0llUxlJEkmBZ7hAjSbDTJ7n90Ke8 ZploTUa2GTTaHuUvg/Wdbfvaortg7KvRjuPNGBHflvJavg3jWzTgEt8yrcnRvgjWNDjam5KJ2N69 uSzZsEmWlMuZMh5riya6gkwZoFg13rhQJeaICoi2Co/BCFkMkbEOxv0CWCcgamCVgcB72I6ohNUY 0leKUlENddWLkQ55YpvCAlGWYS9W2qCkmlHWNul6cTHohiiBcYDQ+Uqtygg6PdJhG973Ii5UiQ7Q /YCJx2DML4Yxj7FhvogTKkSMgPdqOW5Ui3aiXwBltYiaFkU65IltCpXAI1OLdphaIKdsLVoCoxxR DGsOhGqvJo06gH+UoNMjXRVp2jGh8Sl7/UWeDWO08lC9Pt2Ik8+8ws+saNtXVVob5PGMz+Ga439+ e6Yyh1cUPuFcYQSI6nNsj1zO46KMvwP+jRlQx3U+X2P7x7B2UfeHAW8CCpnX1AzrRVxHo8y19Ojd wbllfBP0Vxn3wDVWZrU0m6zraD8lV8XMKSyzQKs/DtgDWMA2qgfczzReZWJRRi0D126LWUaQ5Lln Z9PZdYb2Tsi3gnW/T+NraLqrvG0mPpvhsEsF61xM/GSJvJ05XZ12kGkHTVp7nteN76DGV8/9Ztvn 0tWVj98CXATcAvycfexVblj5+C1+n7cBP2Mfp8sXZ+Njpft7zNfQdFeZ5In4eJDLn5j9x55lVrQ3 mPaGRuvMPDv53hjB18pGe+3jg/gFJ3ATSYiN7GOvstXKxyhjLaATsIF9nC6DnY2Ple5Hma+h6a5y 29naJZ2/13K5zvSLPfOtaOuZtl6jdWbDnXzrR/C1MuTp9JzH39f+UeK4HrvRblt4/oiw3JVc+qlN 5tIVzTq+97hGo/LrimYH32vVaFTOPZ3e+txW4GLrdubVrtnEmZFXtHGmjWu0ziy9k298BF8rc+/V 2HIbZ0cA3wIcB/yU/eJVbl+Ns2dZRjfg2zzO0uX7sxlnz2q6/4ht59wJ8GKcdbOPerhEOc59Ap12 I9NuNGntewdufDdqfPX9hLH6a7qxhLxbuNzNdGq3Qac5wGXSpJE7EDrNMS77TBq5KzHesYS8LvB7 XzDf275nodMOMO2ASWvfx3DjO6Dx1fc2PomxNJ6xV8H2xOvTXBqku30XBGlfAnxNo/0O++aztjOi xvsKHueo8w95vKfbLRl97WLfQdFpL7IcRevcVfk0fIjx8wX+pjkPqOax6NXOi7In9pMqwDmWEaSY 574bk038RN3naHwNTXe1TzPemIk613B/rTH7tn0PR6cdYtohk9a+r+PGd0jjq+/1eO3j7YAzHEe/ xz72ai9I+Xgn4LuAp1lGkNYx7vtD2fgYde/X+Bqa7mrnaLw+Rp2/D3iSS4Pk2XeVdNohph0yae07 TW58hzS++u5TJjpvwnWVj46uiQ8cx7pwt8bck4Io1wGrxS4YyeugTEG5ScSgbZeIZpGnqNpyNeUH bMjzbVNtKv+Bv+l/KaDy+rL/pvq+NBD449m7qcOTCkXowl3aWyK9oB2fX+bvl8kfeKPyS+cDe24W iEvw3EvvFoj4qv5AzcxZtDd12FdINH2/7A+sWhsyNgCwLTLpYKCUZVu85E/Py6izfkh7O34tlcP3 74A8H3x8vf1eEfeoQtF8U+q/Ct7jTt+HqXfh3qXrA4E9h+Qrqnk/4FiTNPP69grgF3wG8VW+d/p3 Uxprw1MbL3MMDXIOyXEij35bMUsOM3kCuOHfXKY3uAyybL0PNPMaezK3l3PM02UXcG49ZwzZq6E3 SMl2mW5yVd8zOP82SeN5hT+OaM8RZusu4NgqOsR4f5n4Ue97+zcXCex7q7jvXflxfyAObQhh6y+f bN8zHDobXEedpS6F4m9/svoe9jnV9ypuufc9g2WofjhWDBAexgCv30n3g+Jj94Pk8/6fC6j8K/Pr K7mbQv+gb68/9c+U7JlC3IL61rz85prVRQL3op1+eBWeR53XAP+8Vf9K9S18n/jgcw+8fDWFz6F+ WOKblIHetxLXaF/bLQbg2Psmr8VRg5suY+8Uj70po4w9nIMbyTcJsMFu8EoUap1Qtpt9P4dlBtQu ghYbjDRjdCL1Zl4b7OS1KOZ6Xna8G35/nWD56d6tmfIdMfi3x4xnCk55cZaHck5yPHXKezGDGNoK 9otBHNsH33E5DplusjOpu+XiFwIeBDRwXqiA6t6cA1DrJswZPcQyZrPu6c4GZLNuqtP4zmMfOk8N jHfd1Mi6NvJ4MKhuP1Ggr3dncznHXKPbTxk4+faN4GudPPB6bbyQ5TXwt7f0sTcnE3Qfn2YZ50wf u59WyNbHpzXdDU13dY5hImvjc1z2m+tu+xkH3W/nuOw3/WY/9+CkHcnXOgvxWchxeFV3riMxzoa5 X+CKqckRA/HeozyvpYuBT9A6shNm/H2AKMzw7eY8qM8VQbWzm2ZNGeY8P35jbXXRY2kGa8o2c03Z HwiMsaYczxwVZ5lhnpvXOfTE/eFFGcwZTaRjTOyFf7kOGxlpbKRkY/1h/n52yl6egew4yG2jOb/L ZpugNt97MV89wmuskJbL8ep8iopl87l/VnNOJ0j5DPczK9nEskc45i9ivoamuzrNMt5YNpPjzud5 vBm052s/6aJo53OJqNT00E+/6HwVrc5XPxHj9XxVzvsF6ONvmD725sSM8jH64nl+r5Ps43SnaLL1 8XH28VnTtvbzNRPxcQf7OMq8nWdvdB+r91PrW+d5HCdfndZ5RsfL+UGtkX/LtvqDyxr5YgYxJ6bl GZwxZ6zYrOvxDu9L3HbR45UM9FhBs1QVRJFaW64joOmjvrkgDvaWa/ZI902s2icB/VzrfE9vDden ae3wPr1f5RyG+h+b5Vr/Kddk3CfUl6akx/r/AUxwvKSEOgAA ------=_NextPart_01C43995.4B4B8F90 Content-Location: file:///C:/0D741EA9/BIBOBCl(correctedt)_files/image008.gif Content-Transfer-Encoding: base64 Content-Type: image/gif R0lGODlh7wALAXcAMSH+GlNvZnR3YXJlOiBNaWNyb3NvZnQgT2ZmaWNlACH5BAEAAAAALAAADwDt APAAhAAAAAAAAAsIBQUIDQ0LCwgFAAAABQgIDQ0ICAUAAA0IBQAFCAsLDQAFCwsICAUICwUABQsF AAAFBQsLCwgICAgFBQUFAAUFCAUICAUFCwUFBQsFBQgICwgIBQgFCAECAwX/ICCOZGmeaKoGauu+ cNyycm3feK7H9O7/rh5wSCwaSwHhcXlTMp/QKDIpra6s2Cwx6dRGqd6wmAcef7vmtJiLVg/L7riX 3Zbr4Pb8mavf1vuAPnSBP3+EhzV0hohXjI45iouPSJNeAlwDT5GVMm0EBXQGB4EICZlFCgsMmpuc MH8CBqufon2lAadEAg0+DqsokXiuwC0KsiKlDzAKuTrMMKXNP5/SNrvDR3/GvwQJyi7UPuEv3dU7 CBAH3QHKEQELoAOXogoBvAAKEglJ5iaS2AC0HQMQaxSAT0naJTEwYd+7XwdBPURYKxmoBuseoqOQ QBXFUeWKCFB1MF2EAQow/1awoLCBgG8nCVyAOAPgq2JsBoYrhSGDCF8hR3xSFi2CslgO9olCp84U gFskARglKCsokKHILgjIdJKpiF0EKoxCx0DBtxf/sAn8dYmowyQaEtw76HQEWbvp6D4ISSBdUK9P 83bbWxcIYAEalPVl8FLEUAQzpfLaQLMmMGGc1o6IIOuu3X28rOKLiq/WUL6m/uYtDbIA4X44Us0q kGEVWLEiZE9VoGEAAZ88/A2yqVkE5wPRTCQTDfipU2qofdcFnBx6YXGus86qwEFD5IJ9kV+o0PCB MYOWAwazGbBFrF/1Qhc4xtvgS9FSSRLgUEAV2eiqGfSJfxAwgF8O9SzE2P833RjQARf67bNAQw7a Q0YwfBCXAkJ0zMVhA7dYiBBpHOKyjigUcaAUf+80FEAtdL2oDigwQhEWekBgOBxAabFHBGSVCUKD jurpaOSRSCZ5pI9ZuGPhG3i0wmOPTObBxmWYVZJllYdc+QeVemzJZR+KtDemcGCeqUWZZqopwpVu ktlKmo/AGaccGI5ApyM73qlGnm/62aefaUjZpppsEupGooeeyaiiYzy6JyKGQjoHFRnqeWellmKR qROTErJep5emh2iopO4gZqBxopoqJIu4CsiQr2axqqatslqrFLfi6iatu57RSavABstEryTIGqav xhqR6U2/MtsslFQqa6X/tNMKCaYhlyQxFwlYoRCuC8kVgm22sKZZRzf6zVfZuCbAm8ItsCVyLrpN 0FlHBCTK9US5qvZgrbHIEoNCMiUUxMSBTSSLb8CqplCPNPVgAAFHCyR1SogamWLRkwRx8Zq59z6M VqhtXEIxhAzQS5dvBfByCwZKISeYU7E0VG9wJZu8AqptTFxCxauF5NVIs6T23GrGvagzybr6/Iqr bSBMQiwUFP1cdhHcU050X+FizNMQ9yz1FG+owBlEn4Co9cZKcaN0JkYXWBrZd1BydhAFB6ECu7MR KOBzkYE798usScXQdTgoMbClzwrigpPsxIgijTOyAXNOKsrIoQZc1Ni4/957Y1mEsvvZBdwcpO/q pCyU8e2EO7XQjqOpPnQ9wlZhRBks4HQNhLtUe8VsVNs8X/XWWWv6E+zaQuF2wjPi1sXw8OyBuqvV u9/uWAGwBXV9I1xqX6vQJdB7ywMOeYQ5PqF3+2KQ5Fdp/qsqn2D04cRTlfQp40vB48wgjAEyAn0l AJBTvDIYxDlwWOU7gQERwT3D0e1w54mIYqzHOLSMqQsTPAQtaMIMBWaiOuBzYAANZj8J/i4BA0mJ XkImHbr1p2V2E18H/dalXmEmhIgo0Vy6RTOGgMIjNXOOt0L0LQ/20E4ujGLprADEQoyqdQ6bIhU3 JJFBvYlTX3yUBI2URf/nabEKg5KXpzClJDRU8YyNMgL17kCkLwoQjrx6wjjoKMYtvfGMVVugNy6i l308QH2GbB/9LuTHNvxRi2hApIzQ0TlZnGd/NRQS3+6IRyh44jlLa4oNR/a1HWajfp08ZfUuyEq+ uMaEaERlKk+3IVC2cmuZLOXO0sbJWapSf7Z8WQO30UCVrTBHdXjkFNGQIhpVUkWggxERYTiBIy6S l730JS1vcMwtytJkavxlDboZy28+LJzO4qZDdsmrL2kzN+yEFo946Ms9So447qzVYAz5PRkhw2OD tEdGVPG6a2JPS/l8VdwyOJX3zKxmTDEacHwxuiklNFWuPOTNNphLvwj/sok2UOYpL0qqjGYQK7oU 5nRAAwkNkbRTrhwACm9JUxJUEILDwNRBFTXMqtzwLilNKQLswzycuqJYLMRoAqI5OH+W6EXPPGJS vNVStURNikqNZzmPilSsltSUvbPqVc2oUH7EQaRbMBtaz7ZWqDnxnWk9qlHh6tY6paWtPsNr3uZK 14gh1F59zVFmQhpYwWqJsIW9Z50alli/Lhaxja0qn5Lq1cgCdrKVxSIgH6fXRLhxp6Xrmzwplc3M bnabjHAkaJeprM4mz7RmYy02u2TO2IZWtKWd1Wpdi0+gpXa3fY0cY0X1Us36UriXDYQhinvcSfF2 tW8N7p6em9vhBha3/7aNlG8bi9042qG7xqUrdqkrHMNGFrfkTZas0mu/HrHXjuY9b9/eC17YSvcf 7K2vfe872rO6F7ri9aGVqgXgAEf3T9sqsIEVjMYE11a+56Jud5lrWfUUiayF0heD71tA/0qWshVW r3B5W18KW7ZSrtWvieU7qBQ7d8PvBNRY14QyGMfYUHrV74zLGGI02WnFmvxCjwW4IxBC0VlvfK+i hiOEKyLZm0P+Ga3ImI1HKhlybUSyMq/cKT6QsY1Kal6UX7seMCdJzGM+mY9bi1xPprm/aENmmzXx 5gMfy4t1DW+deXxnFsx5ryDes579EMaBJVPQD0amnv5c0UAjesdQWv80o9P56P2WrdA65mulvato EU9adnjm9KMHGCUxhhSMoka0oY0cas+GmsukYjORTe3qOcO6y3KWcqsZ6cdNYxhWazgyn33t3P/6 QUy3hly6ICvnQ/t62FNrdLMTvedqTVfL1K4zfj+dbc/a+Lst1HWmV/1tPHFbVGs29JO7rVxa/7ZI 57bubC39RFQT18z4zre+983vfhMpzHntxj3qUdQXlKIWQK63k/PKL4ggYHUyaA67Z7VwtoI14kyb uJyEHVrRCeV9/YRRRjzuaLsmO0wilwg8jjKQhvqU5d6j97M1ITp+OWCjgQGJN7aRczvPfI3MG4nC sHJS1yCt5/T8+Zr/3LW7jOHSOTaUxzEkXl2l8+rgbJnQT3O49Zb5i+oltzqdS3SME6HH7LlJgj5I /mux5xHNbt+qp/pQIq2mEq3sJXja7Y5HvOcBYJJJAUVrMHiCsU4OAk8Yaa4GUhdcY1p+l4PCSMCv SwxcH/xAiDxkFJ/cYGSdkP9XB6mLPMPZozFS8U3hwsISqZweJnQrXLMM6LJBa6GBJJjYjQJTlrMc RiVjsZtZ0DXBA1G3ggNqWWR0F7uvMMg1QAo8AJof+oWNHvExu1rtlMEb36xu96NhwG56cxCIz55D hwyo1zAnyRO+BYkPICSkwzD545REPOTZ+VK+g6J0QIY8GWAWGTR7/ykUIkuRFy5nIIJUQ6XXIDZj EOklPx4iIRTyIPcgPxHSIjDEGCAze28RADRDN/63UdFxNBHCSvPnCODXSQFgN45hS0VHSqnBNV4T TKnmCNF3d4VRgrgENregEzZ4cingJI23TEynACEoTF1XTCYiey+IgjcYd6fDIfAAVTWDdiEDgtTU RSbiTEkkhCd2FRygOhonhYC2A7rzFfUChuelPCJTbmaYXFAWh3QGd3S4BJF3h8dyeHqIh3zYh2Iw Ry8giICIfVplT4VYBR8hSeyTBBPiDYbEiIqUiFaQgJhUSvNzSQtITpT4AwzkDbCUUbDUiVIQg6G4 NTKYSaQYBTN1iuRRN4qrqEdL6A00hBrcVxW0aEwXF4tuKE1JUETVpIFMFTZaaEQPwYtuwInImIy7 uIyIB3rOGI3SOI3UWI3WeI3YmI3auI3c2I3eKHZPlRDZYQOI+I0xcHRPwR8FN4i4YI43gI46QABx wXfuSBAkUXg24AB4U48wcHSUATAcs3h2MTL8eI6hM1UANINFSAAeoID06I7+6JDIkBfwmHtQWJAv cHRAYT00iAKUkxAYmZH9kgs/aFAPFJIqAI/i44R/04wYySGm8T5PBTMCeZIoyU1jiAzmd5NGkIYE 8ZA8GQMZUTlwFQIAOw== ------=_NextPart_01C43995.4B4B8F90 Content-Location: file:///C:/0D741EA9/BIBOBCl(correctedt)_files/image009.emz Content-Transfer-Encoding: base64 Content-Type: image/x-emz H4sIAAAAAAACC9Vce3AW1RW/3yMhgCiN1GFGM8ZWHN8P1Npp/xIpYvlCIoGAtExJ4BPjAEnJhx++ IPLQUEGhBAEJr1hAMMEASUVAHolIqtXio2qtWGWm9fFXpiMy2s7Yc86eu3vu/V67Ic7YZH7Z3bu/ PefsvXfva3+bkFJqJiAEyAdcADtD8YB/Pr1EqWLEL0pGISt6p1LH4XxUWT+YEFHqJ3DuX9apnv1R VbswosCA+rhIqbcAt4+84zag1uM1k36a/wDy9hehh/5g5+S5SuWpGF2dBxgYPg+IuHdOuEd9RXvR 0CA+m32vfxj+hAvgzzfhPBUS4V4DB9fAdkQIglCXAwoYIeYVq/Q/3377rcr143gIUQz5oTBHE6Y9 9HVppFC9rzoBh12cguN/Ak6pLjh+hXGMcIrTnXNHiP83QqcaTPbymdtJaR6OZcyb/BDe5zl0ddSN 5hRtj/nIW+e+9PWlcJ2DLoJOfw/SHHQRTH4Xp3cGKlV7T+doqboR6o2HUnUF4DJVBttyOEaMY+C5 MkaJe81wSB/u5qjDu5HSNMYB/OVoqYgCLfU2R8sY0i6ijCH5Dtc5n93fRYD/hL/kqryCqnwkHCoc Xz0rXlc8Np4sHlczq3K2euHY8eQRQPGRV5NKXc81O8z1GS/D3Bo+oFu9SW1I6dxE7dxEAexHAfn8 5Do+zyPmEHUVpd0xG4gJh2Xy+m4vbOToEfUzwmFGjNIvpbRDaoKLF1SNi1b1KOAxwia13sWTqsXF Q+pF9SBhH+EJ9ZyLjeppFy1gS+OPYFvjJfCpgXHp+oc271ZHGUcIaN8r8S6GU1N0uubrZ+y7yNmz 39PR7lKjIXcRMdiPZeW2wpPUAnhOTYbt5LRcnXstairwEJWEFoa21cLpO9Q9aicgm98dapZ6FrBN zVXbAdm4W6GM/gDYohaqZkC2GDerBqhNDWqDWkbYyNC2NnJ6E9SzDYBsfp9WjWodYDXUzqcA2bi/ h+hWgvdlcEfL4Y6yxbgUcqkBcmgJ5P4Seh52EbQt3Mf0hapNLVK7s/qdr9rVw2qvSkItnge1Oxv3 t+og4IC6F57Je+G5yMaNQ42PQ42fAtspGXoTzZ0IT8RE4JQTur5HT8Iz6lZCM0O3T5i2Rf3KxQZV 5+IpKB+N5VBPlkMqYiGViIO5kNMJwj7CI1D3NbBuaawGGxpNYFtjM/jUwLh0zUikaZ8Son2qgNx1 0Emw2ycnvavXbTv2PoOoT7kS4sKeaUS8DvqePPiNBO4dFkA0iPkML/0LwnyG5Dtc53zQ0m6FntpB KcFra8oYpQSTX8rpZWltOyNdexwaEfmFjGGRMNSpUijbUvf4Kdh/Qhw3wP4j4vhB2E+I49mwf7c4 ngr7dwHSRaXH26l5Xg734aCU4PVsZQzn1+Tr1LLAea7LVpe1l95GWMCQfIfrnA/ur4PRTtDpD0Ca g3aCyW/n9I4suZmuTOvgqnsB+nga7E8GBC2TDkY7wSuTDkY7weS3c3pHr5/ho1DfMKrba2qm9+dR odkbDR/wBnBwlBmfXTN3xj35yvu1mR8ys4bGo/lpx6Pp9grp6glQ5v1gb2S8rnpOfDrMS502JQp/ 81PYVcyeVjM7MadmJnHQXzTwuNacIR6GEj2s7nPRAcd7AB1qDhzXMGoJHZzunJtL/EOEOre9drh1 lOah1ud8RkfTQdvawPOZV6H9cFBH8EblCUYdweTXcXqiF/XKKZ0bYCyUWjoRKp1oSr3pgTzFejOq ekbxzdf2A14+w18vgu0qopnhpX9BaGZIvsN1zvtrV5qhj3bQSPCe0dWMRoLJb+T01b5br8OEBQwv vZWwgCH5Dtc578/LMhg/OmglePeyi+H0kCZf95q7fHopAS4iBr0GQqePhxEIohzGhAjJj1FvM5HO +/UymRCDkRLC81JJKFdVBMlHHqbheb9e4oQY9MAIz8tMQjnMVBCSjzxMw/N+vdQSYjAKR3hekoRy GMEjJB95mIbn/Xp5gBCDEQbC87KQUA5zCoTkIw/TxmeYXaXzsoQQg9kKwvOyjFAOY2aE5CMP0/C8 Xy8rCDGYWyE8L2sI5WotQfKRh2l43q+X9YQYjN8QnpdmQjm0IwjJRx6m4Xm/XrYRYjA7RHheWgnl 8OSVi1mg5mLaeN/P/kRoixAV0C5ViPaqBOItoftbT5B85MUoD5p8e9lNqIARXIUY3ZXA/AURgzlL TMxbNBfTSjLMUtN5aSJUQGwVIuYSyJUSysVtBMmvoPvbRuf9etlOqICrKoS1EogZEaO42wx+BXlu o/P+vGTqR6KQ5qCRYPIbOX21by/pW3hcE3LQSjD5rZy+K0fvfwH150+qgWoA7N06bVq8NlFZNTNe QCO4PO79nZ7dXqNsgqswrZrWKCO9WqM0x3K4EvQMtCoaawmLCVugx3RQT1grzj9DWALXL6EVJT2W Q14zrzAh8NxmSPM3ltPRrCWr9YHHcm9CG+8gSfDsLmI4UZn8JKcv6pPV/r3QO3eocS5aYf7ZoiYR nldjCW1QX9qoznjn2tWdxMfZzR7Y1zmKPLSpgeeezzLHNHNUR4Oe2jLMg7Pl6IvqDsI+NZqg0zFa hI5K8pGHaXg+6Mx0JTypiBXwtCJ0epPaSNAr3ZK/gtYWN9F5f16WA3M5rWJuIuj01dAiIRqhdWoU vYvmYhqe9+flUWAilsBVSwxrrQQ9wpR85GHact991Vr+XcO/npdNBL16K/nIW0Z5sMmnlybKX8z7 jQSdvghiReg1PclHHqYtMtpEnPP8N2y+84zStgdqrvP+ZzCxlDpIa0ahcPhH5ffXJeKz8GiAukT9 ji7HP78+U6hePHo8qd9sOnbC+AqXfn7A94BrDwN5NnUxXHghoL91PICPI9ze6vNfoU3AELY1yOf+ UN7HRneYF0f9FPF+G2/wUm49Yt7b3eIgb36z/Wi/aOF83g/zHFTvVwCmA74GnECfQH6fOX31Nhh9 3MQ+3sbVQMB7/H4/0xviTPl6vrgntHu1iD0acuxGRez63bGfMjvfyhe0fRfHjNt32Lb9XllzzzD3 jODa75ptu2dS7Hrvn4PWuUz7uownAC4EYB28ksu4r95P6zLG+7oCn032gWWc6Z11kDLG2C8WdqMi dv02+2zLeDJvZRnrN92SeyVzZRzy7Xc6rrQr34j7iXkcoDbkxPylJfVYwY0pvSdX1TCbjqs6aC3G wjYJ23GqBtIq1ewA7ca1E48nw4A7C0KdOk23R/gz+L1C2r428uvk6LuWR15//HRyXr8h6pYVp5Na h1MG6Xj91dEiThmi7vl8aWT6u4XqWbhuxRtF0L5eqI4sHaq6txepE9uLiEPbwUXRAsC80BDVdk5B uJh9e7acH9lO5nEack/Vdifz+HwP+AvVK/XQzKFco4ao8e868Y+G++hp+Ca5GM4VfPxjNf0+5xaH pWkr8crxWA7cH3zA2qTP+dzOUYNGFN06aMSbsN/N9T6WqtKhn0lUD+aqBFjDvwXMj/I2n/shWQfG c79xnbD9Q8v3VXxNXg7fd0BtcDybPtP5/S72C9M8g38H3Ag4yVuMy9ZGILcFsExwu7kP/b7pJXR7 iHl8nGN+i/M4k4Yie59j6iok9032o7m21uK7Ls/vYl/f3yuA27icR/P92eqMXDZepvxzbE1iG7Zq I5uNm9jGVMAxQCWXYyY1h/bbyVy8ppr92gqPXLEfwvdFbGse27CVH7ls7AMsALwEaGAbtiIk1/23 A54EvABYxfefSSmi/e5lbgdgDfu11SO5Yt8J2ABoo5gdG7aqJJeNzYDnANs5JrRhq01y3f9avvcm nJnw/WdSoWi/jaghBawDHGa/tjIlV+yPAY5iP8/PANqwFSu5bGC9eU2huJXHyQBbyZLLRjW3M3PE WMZWuOSyMYWvrRJtla18+X9ok7BMfw44wFu8D1sno/uqxYKr+6rvm3bG7qsWi74qk55G50Ull2ml KFNbYyO5mfoqrbvpi3GFHqtN57Iq5vu5zhovNYsxfKbxEtoZQWPqhDvOzAswTko31lnBebZSPEu2 0kdyccFkJW8drqn+SWfX45qKoLPph8fwtsTtQ021kOz3xvC2xO2zTAWRzU2166mKMsV2GWuGI1nW PiJWvZjIz+Be9nuAt2F6Lh39keY087kWwdGaJM1ZxefWC47WKWnOYj63THC0dklzEnzuYcHReibN mc7nZgmO1jhlyiO59pSuLk5hW1NE/tsKKDnWHMPbEndcaqqibG6qXU8p1du6mO75sVVUknuIuYdc rqmsSseVdqXa6mxi7uLty65tU4mluY8z93HBtdVZNjfVrqfYylU3Mj0j89j2It5ivdF6Ls25m8/V Co7WeJ1Nnezi7ctu3TEVYLKedfH2ZbdOmqowm5tq11OK9UWfs4r7mz2AT6w+5xEez+Xqc26HUX+N mk7r0HJ+bq8J1LK/T3hc2pPG3xs+1iPi0NrUwMhiBsxJ8i2fuXz/m8fG+PLJ9v2hD981Yi0k38da SKb98Vxn/0HlqtSAkDPWlDF18lpivywx4RrjSOrzq2G8G6cy7W/1/VH9xixLDFX0Rg3H2qkxVPmI YRrkzGzIkzmwnSn86fyJnsWaTbp3AH9WTqy4FnM/c/pK76fHmB/wHAJ93MfxZ9IABlkfPiHs3s95 Y6sDe7s+/BnHitsk27aVg5p7krknBddWE9rcVLuewvBs2yOsi7t5LRTf69xi1UUct9xwlnUxIupi NEM7MR+foRDO/53vOWUMNdxu5WonRsHzOANG8zerayneiPCf6zlI18/st8aAUR4DSq2k5OKY+oAY W9v6yXR2Pa6pqQw639zA240cp622lH3cBt5udPtDUzljc1PteqrMoGOcEzx+OuGORUy9puTuZ+5+ l2tqONNxpV2p6wwS52a2t5n7zCiNzU3Fp8yj/bw96OanqRGyual2PWVokDgncD2qEGNoWzMqx0z4 rP2Gr4vSsakjte1Krq0tDRrnJD6e7NozVacyzkr2XeXGaSpRbbuSa6tTg8YZ5+MZrj1TtyrjnM2+ a9w4TS2rbVdybX1r0Djn8HHCtWcqX2WcD7Lvh9w4TTWsbVdybYVs0Djn8/EC156pnZVxNrDvpW6c pp7Wtiu5tsY2aJyP8/Ey156pvpVxrmLfq904TUWubVdybZVu0DjX8vE6156p35VxbmbfzW6cpqbX tiu5ts43aJxb+Xiba89UAMs4W9n3826cpirYtiu5tlI4aJx7+Hiva8/UEMs497Pvg26cpq7Ytiu5 ttY4SJyV3A9XiX7YViFnz3tTmSy52u42YVeqlYPGeYjtHXbtmTpm6fs4X9vtxmlqmyVX2+0WdqXe OWicW1Pu21RCZ68jpjrazvsqwbUV00Hj3JNiz9RS23lUIfLe1lfb92SWk6m5DhJnkDFliLkhwbUV 2jY31a6n2g4ap9+xWoi5IcG1Nd42N9Wup/v2Ox9r4LWENTSfceqnnAsN5PesA7LMhfA99K0wI5sG Y4xaGCFUQknPhH2tI8kT8zO5ThHh/XS6kv4cyzB+X2vH1ORDV1ItdCWRPtCVpFuj2MXr6Qe47y2k Pq1vdOx6jQLXspZwmTdw/Jm07UHWKHaxXR17VMSuVe+9XaPo4XFgjxjb2Yp4OZ9s4O1SNw5TJS+5 2u5Sw66nnO9rnSLqJX7Jc72JXMZ9pazXZfwnngO9DhjLZZxJbR+kjPV7Mh17VMSudfi9LeO3WQvz Dq1POrZtjb7mvs5zxr/w/UVJw2Hq9iVX2x0r7Eotf2/fe2zlNn6baONtnb/m7qb3hE5fuJO5tvbf tiu59vcAQdcEdtBz7tmzvxTQ3B3cv+4U/bb99YBtV3LtLwqCxLmWfa8T9uxvDex1ji2i37K/P7Dt Sq79TUKQOFvZ/y4x57C/Vsie9+YXDJKr7e4UduVXDUHiTF/nzO8dbD3PapFH9jcQkqvtHhR25XcR 6d6JQR9bf7loV7T2V/9Xq36i7z3I/1WDvpFQ95NGIU5vh3vz44whsupoh/rX/hYckXrdIXT/aPPZ kV/Tcfc7hfxtwxD12V/R5unkh+86etyGi04ncUyBNl+b+mnSKTmlPoL9qQWDO4dueCyC//vL1v7u g+sx5tj7har/6M+TDcM/Izt4XVHr8SRehzpf2uL7S7iXjxLd9H/EbO1viO9bt8E6HfK/fpj3bUf9 9bx/nkiH8qwfw+Wl/2Pg5Wns4pjoXKXvzuHj/v8AKEPeZwRRAAB= ------=_NextPart_01C43995.4B4B8F90 Content-Location: file:///C:/0D741EA9/BIBOBCl(correctedt)_files/image010.gif Content-Transfer-Encoding: base64 Content-Type: image/gif R0lGODlhHwEbAXcAMSH+GlNvZnR3YXJlOiBNaWNyb3NvZnQgT2ZmaWNlACH5BAEAAAAALAIACQAW AQ8BhAAAAAAAAAgFAA0LCwUIDQ0IBQAFCAsLDQsFAAUAAAUICwUABQgICwAABQgIDQ0ICAsIBQsL CwsICwUICAAFCwgICAUFCAUFCwUFBQ0ICwgFBQAFBQUFAAsFBQsICAECAwX/ICCOZGmeKDoIQUuk cCzPdG3feK7vfO/DBcOBV3j9jsikcslsEoU8hNFJrVqvWGvwMEgoWArAYMFIBBTdhuORaBwgLWh2 Tq/b7cEIS11oRMxufQ5jDgAPCwcSAlN3jY6PkDhbKy+HXAuFK2iYhoiUkaChopGTi52XmQKbhZaf o6+wsU2llZ6clISnrrK8vb41tLpeAH1cpggBfqa/zM3MKy3RDWUBBn8taiJwAROIAMhyzuLjz5zk 5+jpuens7b7Q2e7y8/T19vf4+fr7/P3+/wADChxIsKDBgwgTKkQSYKFDew0fSnQXcaJFctEuanSW caPHXh0/inwVUiCbaNWG/4z8hbKiwE8rKABjtHJOy4KuitGQUtPOTYKuEMgUcXIosgYV3sRR2fNK yxZAWbQYKsaCMQIPwoxJtKwplqcuAQZ1A2DbVDZhROzyagVsVCMFAryAkHbESa1d2VIBC/XlMjZz hw64MCIugbV62x6EKYAs4LJr0kI43HVdYiZhA56UVkhE3AAYCGTQMFUEOGPxLi/JLJFnjAEaOqtW wlofspRdkmXyAqZsiw0cFCQQkjsbbNmzGQJEEAYtJwjKdBdT9CXZkFzQucROTvufpcJQulCudTVn +ASHt3NX7u+B1REQwqv61GqRq/jX56tf/6P2vUOy6YSLKfWN5xlZuOzHX/8P/t2DQHgVJCBTgbrY t0wXE3qiYIMLwsBhPdBYdxJZR1EjxG0uEAXIEEfJ9mGHJ7zolYwwkkBjTzfWCECOK/EIo48iAbmg kB4RuZ6RGiGZnJLsMGmCk5dBiZGUI1DJlpUcYYllU1syo2WXNYHpy5c62iAmSE2cOZKasmhZppkH ufkmDWzG8qGPdRYZpwwc9mXjnHTuGUOfmeW5kaGj3Okha4he1Ggoiqbg55+A8imoDpNSWqmHBmVq pn+PThQqJJ7WUKoIoz5E6FMx8tUqq0+6Gqtbs/40qEunoropChHZ+k9DjHaEkq6SppoQVMY2AuwM tPKaLEIlCbTsrbDyuuv/op3uyKms1l5bbEHTWttst96+SlC4Me7oK7nllgDsszZpW+y67LZbJarw fiWvs8Mya++T9waEbq005gtttqD26+8IIXJGg3g1fKcquAPbmGu9ZQnGAk0/sEHWxOdWHPANmUFA 1TcfJyGxQwZn0aukOHhqMgmPGWKGUclUEMEtqgx3QFFimGHdQsiCS+wRrM3M8DADWEVJVmIswIEa 8FzntCqmZEd0tAC9zF4JSqs1jFkBUICWWjwbQXZKp7B88T5eI31C2DbPVZeKZxwnxjJ0wQfFygpx 7U/cPpRKNwJkKT1YYWdsd59gPretastOHV04CopvrOJckWmjgON/oVdW/4TNtcFU4G/f05eTmTWM zemfhTHaVEcdcBtcLYTxmW4SvSvwyJf/m666XQPvLpzCD79vPxWBinzyxw/UPMymQg+w9MZnj631 mnY9vfIL22GZqNhrv7ylShQhwwqpuX3+4N1XeSOVrr2m4Napqx6/5XxyeBtxgNhNdcKwDeL0zBoB FMP98Fc+/jmQUzBgjs0U8Jzo8MENinhB1bDjBr1ZRHBw21/MYLCyLQTNQBTa2wtMKB4PfjB/85gU khrkHqbgZ29oIJAnLpQ1+aBhgW6DIUWit4OLAagwCLIQeVzxCQHZB4hEUxflVkNETMngQdeJUIau UiEjwERCuoBi4PA1xf8khMVIHwqRG0bEomSYyHYpuZ3NhtYiR5kPIlXMQRnH+Lv4CXGPfPzVGfM4 KO5hLIR5nN8hd0U45l3vjo+UnxB/VLTiVfGPklRYuzTpyEuaii8gZOQkh+jJ/oEylKIEJEMKJcJu gVKKr5wTKunBKPkFyl302uQs5VHLEfLLeu9SJQ9KxSNhRlGK+qhNMQ1ZyAfScpHfWs0ogzQ4ZZJs lbvUizEZRL3qDZOT/GlkPpznzZiBk5Lvw4fzZDS/c5bJnfVoUDbvBM9KZbMd8mQnv6b5zm0WcXu3 stg9y8VPjjTTUrlkJjLH6UpT+jOKBWXJq4g5rIeyDF/6w6Wt1mXRYxL/L54CPSU0FfrIjjbzlJEi KfrUGcxYRlOlB9UfSt9m0oOlM4bhGylMK6m6F8kTpqZkaE5xCdT+CXWlsSqqUTNKLmXWNIh47GY6 I2pI3z1TqsFS6sKomigIZo+r3GveUyE5SItp9ZZXxRiuxtq7qKrVrGcdKj5jejSwMpOtrYSgXYHp zHPgCaMKgVgWxjc+kokMHX/dqzzicrIjqM8EaZAN+5BjTrx+1LCWPUIHbpiE+pWAsGKkk2WBNNB+ POACRyTB/1TyPw/cLGrUIOBSvrGUMUToDApkxWvNSRHSZvZTMw2ucIdL3OIa97jITa5xHwKbIXAW cBIjGwVyoQnEEIK6//pxgHRHiCzFugyc+XzIZoZmiPfYxbx9YxjPcsiIuFDNHMw5TnqtCEuMkBGu 5nIIBDoj2NQSxRyKuwB2DRS06WJiwPIVDGGuudBxiLVXCbspUMz7DapgUQwM+AYUDAOZLuiug3wj zoE5MYnYPGa/lTVpody5uoVAYyifkc/r9oaNbWhFaPE4TXGmpt3Ztih2evRuHVacqxYvqLBDFvKQ CeYsSOoFyTYpLSzWCuEV5/XJe6BsvKbqpUwS1ZZXjmtSrSzRBn85Wr+d3BmFmdBMslLMn8IvSOop SfAhRLAxhHAz6AzmSM7msaBoM0n0vChycsezpLrsmNb6vnqmmQfU5f8NbscQ28h6TClsi4SgRZEp NM/y0TqYLDwc0IdrBOK9rEBEBjnNU154KiSblvBABqyL9eYihZpWshMoepP8gTrUoFuirTmBa1Jt 89VVVrRXE6I3+tgiFYch9g45dgcpOwLZsBzor3PQbB1yYRjF+MRR9EDtJMua1QQjNFoTUrsSmQGB 2OjMNrrRxnBU24G6/mawHvw8j0DZTgOzdhU0WZJpblsJ/wY4lRMVzPrWmcH+znKWHg6pB3eanwc/ VsMZ7mWB0hfOnxT4wC3uJ5GfG+QaNdjqKtpdImUcWnyOcrNiTVfvKffmOM+5zncu0uChnFqD5jfF ufnzQucL1iX37sv/pddqYwvdzP0pul6TmmQZ3peKUqdrvgPa6HwvPVtXp0PJPb7rrJ9069sLl8nX bXYmT7Ra6QYvrdDu5Kyv/ZuAjXnU255fcyOdTV9neqNaHnZ98T2kVV+o3r92eHnlSVhXf3zj/byX NZfVZZMnZJosbmfFZL7uhXt65wf+edAXEe5SrVzpT44p1C9b9aW3Es01v5fVsx64BZX86ll3d93P +Uqh36vvXb14jxrdpXvH/KJdb8fiel75c0b+C3lO/epb//qsCm5cXZKbsgHaBoALc6Clr9WKnI0Y csHBpXU6/tmTNCIxodndIuaN0Qe97RERhAoSGDTe9Z+8lGd7mqYN/ylDAhyECmUBYnNRgOIngEuW gPwlFRTAQuhBgQRgQuFneg6ofBwGHzJxQ5oAgqpwQxnYVxvoE2ITDjPjRBeYRARAgvUXgCd4b56R aTODIcKwRWezfjL4CN3HgHZngELjf2yENx+zO+3TgHTQgR1oAt83A094LcMXCR5GAlpjAoh2ReVm T2I3CvpnFwF0aQV0CZXWBmtgOmO4SV0oClcoNt2QalyhQSsiCLemannhLVMICW0YNG/YRWiTCqvg h/aSh4/whSpCQXDIRMOWiHcohVcAMflShfDhB9LGRVEDbXVoiWpYBdDwAgbDhHIhbslAbmLzYVw0 bo3ISI9oCi3zg/8qMW/eoGM4Jm8tQG+0ZW+qiDeQM0BioxuWZjpzNBWa0BvtkHBhpRZXQxnS0UEV hICtIDpCsTdqkGnqYA5AFRHb5WwHYIGZ+B3BEAAxeA6jVlTYOH/aKIKBSCE3yIrhOIM0GGAqdAos mIlzJB+e2I7u+IAnBm056IfjRmHxCI6nk48PiH64RVvT8G4/syLaUIs7hBIEUCKzRZAoiHAZRhQL ZoIU+XxIIEGmwQiBBz1U0n3pp4QbiRlreJIFaXgqmZIs2ZLQ95IwCXsyOZOkZwNRCAQgaZMxOQNZ GEE7yZOwVxyDYIZiGAc7U4ZqcJSZFpLJA3/NmIkZVBzbiGqnsGr/GiiUw+QZ5xFtjKhBiyiIGqmV e4eOmQgTYVlsTik8+eeCZ2kKCFaJa7GW/wJ/YEQhqCiHpiiKyhCUZFl2RviKDllv1+B/sFhvKkGX gxhq1lhOf4mS3NaYbPeYZsRtEtdvlFmZW5aZnNmZnvmZoBmaojmapFmapvmYruMCeAYDaXiaTvBc 8wcDi+OaVsBZN/AAWkabS3BDHXAdxDaEJwMNuKibSWAWIEY10NhYI4A4uUmcPsCbvrkG3mBCc6Oc zvmc9nYcn0A3JVBD11mc2bkdJzGcdpGR3/kDtplbTTOQJ9ABzXmeN6BGsoGKEEkzQuNFAkCe8OkD A3CRhmCe+0kHDh75DVsYoJw4hAW6ICEAADs= ------=_NextPart_01C43995.4B4B8F90 Content-Location: file:///C:/0D741EA9/BIBOBCl(correctedt)_files/image011.wmz Content-Transfer-Encoding: base64 Content-Type: image/x-wmz H4sIAAAAAAACC6VZD2xW1RV/795333fv+z4oUNdgShsgFCxadS5rbKTEMWEFWldgEJsCgTkxJauo haCTJug0aNIahn8iFgkOyUpgwaAgGGIKseNPNGyQdeAwRSMZazQMsaadZt35ndfzfR+PZcOszY/z fuf87rnvnnu++70+/vLhH171vIB+R2eeHLnMw88nC33PeZ5uzSDyKvs0oZSY5ynf9wzZtMIoXGVU 68gRGcRSvvZSZIuCJ0eOzgSc1/NCGhGPVF4F2Rk+st1EsMPw6Rc/473//DM0NOT9rx/MUEFpcEc0 EU2HK6dorMIc/yQez+KpWOt7s8hO0hO8NVFBakU0aCZH/cEBVxhUuXt0m31ZfZgqVV7qqD8x3OfP NEf9tUGpOqhfVml9j25XhUGJ6g96/UFz0i9I9fqTbIma4tqVi9L6bHRQr0mvDbzMTDMzMzF8MOOl ZmY+THmZNrsmXeXORgeciyZHU9yKaJLF/HGl/p8KoUb/DfGqFdt/fKdKecNjfO6GSXruNRU74Mqo agupap1UtduoameoakepameoarepDYTjulMV64V6iyoLtlD1Kqh6l6h6vVS9S1S9CqreFqpeMVXv OFVvA1Xvu1TwGcIauq4k3zGKHXPF0WzXF3XYDemPU2No7BjKsYhyPU45F1HuMZmDekM6rfuidlUc Yeewg9jJeEcHDXYYO40dx86jA9AJ6Ah0BjrkY0IHdcts6ppjVItKqskaqk18PwWpiGumvdw95nwz woLUvHB8ClZ8+6I7ve5oNtuR7DNZnfjBSzk2wns+dVI/n9qtb7HAe3prqjDYmtKB8OdT73F8KwPa k1ryQgcO/+/pGlw+s6LJzy966dar7w84Hs5gFKQkNi9cPvw7aOYxloe52KCB/5Gwh2wPxyX2SLje PEl4PewPfkMAlxqBi38U+0KvK9wa/DF8MxiVqgk8RmUAn+QDh38Uo5LjEltK63qI63ae6wM+nmMj vVfcVP9Z97VXGX3trYgukLV+p3uYsNSHD/xZZ1nzCqGT9BgjuaEDh/8AXYMna5w/h+iTNV5nO+jz BSzU6xgd2X1ss19QDwInVBvjC5WLnWD/u/Z+tYsALrETNqM+ItzoXvAjArjEGlyd38T3Mo3ubaoP Lj13zJWqM26+Knf3E+arA+4Mre3vvvAz7m6OH+NYqYJe8kIHDn8nXYMn65GfX/TJeiyiWs3i+tf5 swjg0gtF0Rk/ZHxDyKgisvDJWPgQO8b3dpsCl7XtsoXBu7YmGGEbCDVBhz2vO+yXWvi7tpLjuzhW GEAveaEDh7+NrsGTa8vPL/rk2ppo3Q2MTtXEuF/J2hrcKD2bUOUeI/sc4TENn4yFD7EbXYcud7tZ I7E+q4O+4Tl32bIAXPq81t6jp9kJ+iPqncidUH32X2qdfU7/0j7GPvBa+wPW1NpRejFdY4zkhg4c /nX2QR6XXHv+HKKXtctnG3Up4Vo3ZD/b4iuh86zEnqS9eI/3Q8bU0udhMc0JK3ONGNaKHzx37sZn Lqz4Ag97dV7D5s61WCd+cKmXDZYaG2wyK/VScxdhpa4lPpZQzr6Yr2fNgAagX5o936ADh39Al/O4 ZL3y5xB9slfGkm9ssMrcR3PdRxZc6vYO+d6h8f3BdmMM7CYDn4yNfdvNTNNjFppu1sjYF82gedFE 4RuEt+n6Dea58zn2RaEOW8IxdKa/nRfDd8Adw98D87jOx0Op2zK/LFjmNwRHvLLgBcIRr5D4eb3M /1IfyfIa1tQzoC/LfsagA69nfMnjknXLn0P0ybrFmspgG821jWz+HEPkG6LxFWoloSEAlxg4/M2q nbCO4xL7QG0NDqk3gw/U4eAc29z3z7lhf7FOmZt1P8cl1ki9MpfQqLHXmwy41KtRR2Gj7qEeicKb gigc0OPCubqeMD2ED7xRXzTQrGRAH2W/Y6EDj/3TeVyyXvlziD5Zr7k8bxSm9aBJs25cOIZjKe+Q GjSH1EVTrXpMHaGacdHAL+MlVqe6KdbNGold8t8yl/zt5mF/FWG9AZc+vN2/y9ztl5vL3ungE8Zn AXwyFr7LjMMUO8waiR3xGoI/eXEfvMW91ZA971rp2a5VldPz52fBHkKvf5r4m4TDQW+WjzXQNDOg 78/mhg68mXGYxyXrmj+H6JPnXaxJGVipp/iK6XNfTL1wMwEWfhmHvZY9l/lEJ35w0ctz5ry8M+/X 3o/SLd7tadjcM1msE/+8vDOvJrhC31xT3E5t3K8IOzX4HkKX3ZnlRQ6aOxngV6zkhg78TkYXj0vW LH8O0Sd7MT5TClJ3hAvILuAzRmITQ3rSDw+l3jc/tKfNTyy49NJq02Z/YZ6w3zN7CF22jACfjIUP sX6atz84xRqJbaD72BB8auP7Mw5cnhm2+X3RNr80fdnriy57Z6ON/g7C/kj4Nn9EGvGNDGj7IskL HTj86+kaPFmT/PyiT9bkLNXrOOGgnkN2jgOX2E79Z7dZv+/u0tXRXP3TCDx3Lj0VnVOPRq1qB2F/ BJ7r3/3sryNbp45yXGK9dA+9/tkovrdvI/DcmdVm5+r77IB+wt4UPEG2zTbqU4Qu9g1wvIk1jYSV HM/tA3TgKxldPO7aMys3h+iTNfnEwx7sj3z/UbKPRuDSC6N9F01gDDjY0WThk7Gjh/3L/L2u3n/d gcvYjf5it40xh+1GsvDl9jT2D/lF7pJvOJ47B0/ZavWpXcLA9ansusHhP6S6CHs4LrFi/Zot1nto za8R2iy49F+rmuN2q8WuWv3NVRCq6e/YOoLwLRTbzWh2rWShl7zQgcNfp6p43LXnWC6/6K89x6pc O+epytYJ17uVIRS5QwSJS17EPqf1fq6uWGjAJd9m2tPN1DOwohet+MFFX+6Oh+XuryGs+E6lN5lT 6e0GVnKITvzg0rcXou3mQrTKPEX+vYSn0uAXCT3si/l61nzFAN+ezQ0d+FeMHh6XrGX+HKJP9i3u qYow271E9iVeU+7vsvqwyS0PD7hxYaeLQnBZL7j4RT/g6HvdXTSTaa7JUbcBlxg4/IvILore4rjE dhDfQfwCWdwjuPTbvqgl7I5+FxZFTxNawn1RBWF6lndHL3F8H2EHx1uy9wMd+A7GdB6XrFF+ftEn ++23VL9NBFjpN/EtSI8LF6Qrwp+ThS/3buVpzgUrc0EHvfjB5U0n7K3DbzxvUMGItcENdPWN+v7Q fH5BtpFflAXKG/WzpuYHWsbf+8Da8fNXNS9/yJN3jPG7NaRz9M8EE3qzmh4cX3krcsVZG716X7LW X1dWnX0fJzl/vGb1qhUrkCnOOc3P5bz3Ou80zqmzOR9e1fTQ6hZkkre3ufeE0gfTbCf9TfiM+sie U/ib7hbGuSyvtS9zfDGhlrTQS92hi8fDf47HJfsgP7/ok32wWU/1N+tpPqz4pnHOg1fNJzrxg8vn vk6V0fNWZVCgTtOzfsoM0bNfAb/hLRz2naZnvErWtKqaoI4sxkhu6MDhLyELnlxL/hyiT65lJ9/j VH9n3lqWcN6yYEnefKITP7jsyRI6H9vpHK1We+ns3+tKiFcQhLdSrJ1R5JaQhT73t80Vu4RhSAtc ueZ79+r8sf561tE8rG3Om0904geP30En3+zr7Htr/A/Dt0r+V+DqN9SK3wB9o8blvxznzl/weMvq B5qTb8Hjrp/iDw1plv4baZHj1hAZAAA= ------=_NextPart_01C43995.4B4B8F90 Content-Location: file:///C:/0D741EA9/BIBOBCl(correctedt)_files/image012.gif Content-Transfer-Encoding: base64 Content-Type: image/gif R0lGODlh+gAqAXcAMSH+GlNvZnR3YXJlOiBNaWNyb3NvZnQgT2ZmaWNlACH5BAEAAAAALAAABQD1 ABkBhAAAAAAAAA0IBQAABQsLDQAFCw0LCwUFCwgFCAAFCA0ICAUAAAUICwgFAAgIDQUFCAsFAAsI BQsFBQsLCwsICwUIDQsICAUFBQAFBQUABQsIDQgICwgFBf///wECAwECAwX/ICCOZGmeaKqubOu+ cCzPNBvUeK7vfO//ssANSCwaj8jkSqhsOp/QaIkprVqvWJhwmE0KhAMC4BsId89XKnq8LRgOKgNC vBIkCIoFw47Xr/9OamcGDQMOIl93KQoPdCoQBSSQgJRNW38RhiSMjiYQiiuTI6KVpUWCXQYLoCMS EwsBkRABdxFtIrZCDCJ5bb1CkabCPKhdX7snhMERdxTMdM9swcoj1MPXOZeYAcgm1tHReRUjmYfW AOfY6jHFxtwo34rRApqIAePn6ev7S1x/ecG8NVgmT9GXcbwW4BtYjSG/hym0UYJQr9UhUeAUARxB T0w+hxBLbRlJsqQpVfdE/xBC+KlPGQLPKACYtRDZx4AhKZXcOfLaLDCHEtLKJKYXwlwp0cEq4wAl 05yVePKESrWqC6kmrWrdSgIrACb+uOpoJ9bS1Ethy9boqVYJWpIjqKRtG2SIRLo+2MIVQRYvO1R3 /WbjcjewVnE7DPPtK3gG4LlWEaeQ0+kE24iKG+NgLJhTP85xM2v+O/pESxWXX6QurRpyKVUMbFVE 2g2xMqMAkEYiU8ACndVaRIt1bZk4JZQXGBAyMy8Wm3soMTRq6UyRgUZjQAF3LLyq8SnfAVkTJzm3 JsmnMxZVKML31xugg8fnFz70Q0I29dALymYceoJiRDOTLvbN95eB19T3Hv9E+KmkX1IAIPYfLgU5 QshLcinIHVYcdqjhWMEx2AAyCmTgwEaImDFhbhWio0FCFYD1oXw8dWDjjTjmeGOHbrUW0nJinEYT OiPCKAJGFVJggIlEZkjESDqSpOOUVOK41yk+TqEOIRfAwgpSCP1SwU8F5HKHUQZQsFSMds2IGpRV diBlnHTuuB2IV6XlZhUNJqEGgm9uUaeVcA5a5ZWJtabnlkX6SdieoQlqKJWFThrlnezk+Rij3Rzx J6RzWhpnqKLaiKgRWXW3hlM4PemPm6SWWmesop7aA4+stfBpXULI6iutk9o6mFS55tmVhsD6ammy s6qq6FTF6vqqgswqW2v/tZQ6+xmx0fYDnrSSWitujiVdC6qH3Xp7rA3hjusuudgSCuhiwqaLmQnG Vfruvpe2OypZ382b7qaY9crvwYfGa1i+AndLML7+IixxvxHLq2Wg9ob4LcQBTOxxs/qaKqdc3kKa azGPdfzxyiC3azBYJWdsLMTHsmyzoRyOTFzDDs817c1A41xjXzxH286nQSctNLQ0y5zlxgYrLXW2 Xm3stLQoyDj11lOiS+/VT4MXNddkewhz0UZDVjHZbCesLdhEq9z23FRjCnbTENOtt7w73Y11cXLv zbbZ6/pduNiBCy612aY+arh9eY+t+OJSjWznYo9jDl7jk3NdueQyZo6y/9ySdx50zv1mvuAUVpru ebmlN676Y3a6PvVOlif8OMqE2q50r1DGTujuYY0tvO8eA/9y4uSO7DfvvSN/s/G5Zys73MUnfrz0 Bxu/vaRoD5d9lNzbXPrxLpssPuvkl28t4/CWGr5VtMfvvrLwty6/+loB1jXz9wtWhy71K/5552cE DKCsBmi/Bc4vJ4IQ3vYU6DbcMW+CdHogd8ZisggeioL7q5Gc7AdAoVXBbgXjWQRLOEIQCrBjpArc 8vCnQV5RKysbjEvCXLi0xoGud/pi4aWwYLeq1aVm1uPhoEgHQCbqDHxCTB0RA+M1dh3ue1FUYrB8 WLn9nUEvVdyWnjKYRf8tLpGLuCsgGnryKFwBziRZLKP75Ni1J4rQgX8ADg7j1rc4mnFHNGSgGvOo GG2dpVl/rF4IUTdIQKAwUHuJIx2lh8EdfqyGP3hk1uC4xEn6rpJJvKQBPfW2SLGxk38E5f/MN0pS IgguoPSk7VSZQJZhElWG1AvOUknD07XSUaDR5S61SMsGsnIfRfyT0Igpy9ol7ZbAVJtcqBa9RF6L co4El57a2EQnWvOa2FzDDVeYxgQ2M5HF5Bc0N3MgQRLwnLx8nThn5KF/wZOZZftlYmCFFTJ+U4Bt W+da+HlHalLwnvdUpz73OdBy1lKB8Exn8hbKQRrAEE6xS+jczinRier/BFZ2itoPwUk3ju5NoIMJ ArxESrplre12zeyoR+lXt9qV0aGDk6VMRUnTCjqzgvE63QInh9InSFJ5IkUjTsu2P6JSVApynFNL 3ZlPkgquqJYw10VFFtJy2RGmtXIdZ56qsawtsoXni9VO3zfJtdqSYVhNoVnD+kRFrjSAbn2rGEuZ F+gBNKTdC6ooZwnXMLpSh1ata2BfurK8HjOFHNKco77S1OYxtoCOdSkldxbZ9/B1LPJbpc4QFrLB Ii9lbtQkEGR4xjoCVqGZ9adY60WvK6l2tc2bVSh9GFhWEpaJkDUlWXVV2Ye+NqDL7Jy/+Cjcr6Rh peWUYMUuOz1UzjE8/5cZLnGB+lPoRoy6jW3tdbd1sRMOtW5ePa67wHtTjcJUruU1b2jRaEezube7 lEIkRLOXGe3a4Lz/Y6R6xTVBP56Ujp8SjX//O98AF7Rf7/qebEvaVvh0J65LKC7FSmtQArNQiLFd rydp+7Us6LSE6TXXuCQ4qgNrdXUcW/B2l+U2ESc0o5bcKIK5dSwMX6XBb0XoBXeb06Xx2LMy1gKQ WRnRqN6PcEiOypKnF2JjftK+t4VqYn15XyvPFqlujMqIu4zZ7p22UFA2RTHJzNb7Vvl9DgYx00Qy Yr2B918nhe6yoiyMNV9Vpmxeb1flh8w6K+7O1SypYgm9Dj8rF30s/f8zortWaM1+OZSBtvGKK73F M19QZW8uKaeHuVkZbvWfchq1eLkXLhii2lSq1i1ekfrqVKvD0fsN9UljjWcQ6nrXjd6xr39tZ15/ MNfEFnWwO/1kUGf60MbO73WfLdZou5bV1J6ltb0821qPatuJvrS3DwVuQJ553BksN1fFjW5yL3vV Tm03GelD6kMnG9v0hredsz3rfE/YxfLupL9bLOl6e9vHt9J3kZOLboQTQ+FbQ/S9DzzwYwd0zAF3 eGIg/sw189upFRft6zTcbo2DNoML3/K4TZ4NWcM0ph8H+UOiGHMivzDg1wt5bjvO0YnbGyI0/13N v8nyk1vc5YHEuUL/Q/Jhl04av0r3MASP3rIU5zjqIp66tF+Yv6djfd5Mp3rV+wn1v359eFpvH9e7 yFsH+vzKVEFxy+qr1p57/dVFT7ja1z5dmN+91nl/uNorafW2q5LDZ3de3MmHeHlZsvHhTjy8erpU 9IIOzG/He5KRMGTLcxjzf5f85L2zd74VWJCZT+XmOf/5w2M59czcShchb3qyi56GUEkw7FzvxAff /loQ1H3wXG9l2v9edzMvjIBrvPPjR3j1tzoa23tdeudLnT4AkxIGXzp0nAd+Q5CMZfeP/31X+b7s 1pdY+c1f+dCnH/hlmT76309a6Adiqe6nv0vbgv/xp3/9p1B3/vd///Z3f5UCewSITDbkTfoHNAD4 N/JBWfnXgPCHDQKDNBToS41mUfTCVhm4f7fWQXZRZh9oQhYogoilYiUocCfYUOxzcyvIgnhxNCoX g0M0g9lndjb4LzgYKJ00gAf3gJkUTEe1gyCjFgaCRUAYhAU4Wa2BckYIdlzxWQh4dp81DCq0hFF3 hWpWNFW4hU1oUZy1T1FoLoW2SQY0gR8ohBB4OJvRJmVoggqINwO1OXFoTzMnNimVN3eIfLnXRs/y Jn34P3EnIyTmGIN4g0zXWdGXiJeziGEmeInIhjTyehykhegUhhUlXHNGhnFIiZqChln2F2UIim2I MbhkhKb4GYRkg/+r+CZiVoJcOEWa+BkUOIuq04H0h4u5qIvWx4u96Iu/B4zBKIySR4zFaIxYN4rJ uC1WiIzNaErL+IqaoYbNVotXY3zMRI0n84VcFo1G5Y1MBo5QoI3XSI5QJY71h43NaI06xo7RaI6m A43oOCzDVo/idFDcWIzymE/wiI8F82j/CJDadHH7SJABKXQDiZCXGGQLyZB91Y9x9pAQGU3bd4gV 2Y1GlJEc2ZEe+ZEgGZIiOZIkWZImeZIomZIquZIs2ZIu+ZIwGZMyOZM0+QTlUZNbcSEkcR7sAQMo wQo4SQn7QQ78EQOz0CpBKZQVUQMUUZRJWQlDaQAbMAKScSFLyRGiEPKUpTCUCgAHz8ESuwEMJEAI CdAAQgCUWokGZCCWRqIKNVECeXABh4AiabkGUemVEcIesEEkrVIi/FEOdWmXmiCVVNmTz7CXm8Ak KeKUgXkFQwmXPckqAtENHdGYZ/CYm6CXHMCYJHAQSoEQlnkFOukchSkEFTCaL2ECa4kQqmAGoVkK EuAICtAprxkSAtAps1mbVAGY6DAHurmbW3CVHBkCADs= ------=_NextPart_01C43995.4B4B8F90 Content-Location: file:///C:/0D741EA9/BIBOBCl(correctedt)_files/header.htm Content-Transfer-Encoding: quoted-printable Content-Type: text/html; charset="us-ascii"

7

------=_NextPart_01C43995.4B4B8F90 Content-Location: file:///C:/0D741EA9/BIBOBCl(correctedt)_files/filelist.xml Content-Transfer-Encoding: quoted-printable Content-Type: text/xml; charset="utf-8" ------=_NextPart_01C43995.4B4B8F90--