I i I I. f., !,' `1 I d kjlll!!h I. ,I 1j 1, I f I I. '3iNwaanoJ, / Ast

Transcription

1 *1 '3iNwaanoJ, / Ast :. i) y ii '' f., w, f!o `1 d kjlll!!h i..,!,' N, 1j 1,

2 THE NSTTUTE OF PAPER CHEMSTRY Appleton, Wisconsin COMPARSON OF OLD- AND NEW-TYPE BURSTNG-STRENGTH DAPHRAGMS Project Progress Report 155 to FOURDRNER KRAFT BOARD NSTTUTE, NC. March 16, 196

3 TABLE OF CONTENTS Page SUMMARY 1 NTRODUCTON 3 MATERALS 4 PROCEDURE Part 1. Part 2. Part 3. Tests on First Lot of Diaphragms Comparison of Bursting-Strength Readings on 42-lb. Kraft Samples Using Old- and New-Type Diaphragms during October and November of 1958 Control Chart Evaluation of "New"-Style Diaphragms and Comparative Tests on Kraft Liner during February, DSCUSSON OF RESULTS Part 1. Tests on First Lot of Diaphragms Part 2. Comparative Tests on 42-lb. Kraft Liner using Oldand New-Type Diaphragms during November and December, Part 3. Control Chart Evaluation of "New"-Style Diaphragms and Comparative Tests on Kraft Liner during February, LTERATURE CTED 2 APPENDX A 21

4 THE NSTTUTE OF PAPER CHEMSTRY Appleton, Wisconsin COMPARSON OF OLD- AND NEW-TYPE BURSTNG-STRENGTH DAPHRAGMS SUMMARY Because it is no longer possible to purchase bursting-strength diaphragms of the type in past use, it is necessary to use the diaphragms currently being supplied by the B. F. Perkins and Son, nc. n anticipation of this changeover, diaphragms of the new type were purchased in 1958 and two studies were performed at that time to check their performance. Because the nstitute supply of the old-type diaphragms will be exhausted in the near future, it is planned to install the new-type diaphragms in one of the nstitute testers beginning with March 1, 196. When the supply of old-style diaphragms is totally exhausted, all bursting-strength tests will be performed using the new-style diaphragms. Beginning in March, therefore, Project bursting-strenfth results will be based in part on the new diaphragms. To prepare for the changeover, an additional study of the performance of the new-type diaphragms was performed in February of this year. The results obtained are summarized herein. The earlier studies are summarized in Parts 1 and 2. n Part 1, six of the new-type diaphragms were studied using the samples of pouncing paper and 2-lb. series combined board in use in the nstitute calibration program. On pouncing paper, the new diaphragms gave test results which were nearly equal to those obtained with the old-style diaphragms. On combined board, three of the diaphragms appeared to give test results which were well below the control chart average for this material. The

5 Fourdrinier Kraft Board nstitute, nc. Page 2 Project Progress Report 155 remaining three diaphragms gave results nearly equal to but slightly lower than the control chart averages based on the old-style diaphragms. t was also observed that diaphragm pressure appeared to drop below p.s.i. g. rather quickly, which means that more frequent diaphragm changes must be made to maintain past nstitute standards-- to 45 p.s.i. g. at 1.8 cm. n Part 2, samples of 42-lb. kraft liner were evaluated using old- and new-style diaphragms. n general, the two types of diaphragms gave closely comparable test values. Before initiating use of the new-style diaphragms on Project samples, two additional studies were made. First, a start was made toward obtaining control chart data on each bursting-strength tester with the new-type diaphragms--the data to be used in the nstitute calibration program. With one of the testers, the new-style diaphragms gave results slightly lower on the average than those obtained with the old-style diaphragms. On the other tester, old- and new-type diaphragms gave closely equivalent results on the average. t also appears that more frequent diaphragm changes will be required--thus, confirming the observation in Part 1. As the second step, all liner samples submitted during February, 196 were evaluated for bursting strength using both types of diaphragms. On an over-all basis, the new-style diaphragms in this phase of the study gave results about 3 p.s.i. g. higher than the old-type diaphragms. The over-all increase of 3 p.s.i. g. was somewhat surprising in view of the other results and, if borne out in future testing, would produce an appreciable change in the apparent bursting strength level maintained by the industry.

6 Fourdrinier Kraft Board nstitute, nc. Page 3 Project Progress Report 155 NTRODUCTON Jumbo Mullen bursting-strength diaphragms purchased from the B. F. Perkins and Son, nc. in 1958, appeared to differ in formulation from those in present use at the nstitute. Because a change in type of diaphragm could result in a shift in test level or cause other difficulties in the use and calibration of bursting-strength testers, it was felt that use of the new-type diaphragms should be preceded by an examination designed to check their uniformity and test performance. For the above purpose, a few of the new-type diaphragms were purchased and subjected to a series of tests simulating normal use. On the basis of the results obtained in the first phase, additional new-type diaphragms were purchased and subjected to further tests in preparation for a changeover to the new-type diaphragm when the stock of old diaphragms was exhausted. n general, service life of a diaphragm is limited by either (1) rupture, or (2) specifications on diaphragm pressure. Thus, nstitute Method 96 specifies that the diaphragm pressure shall be 42.5 t 2.5 p.s.i. gage at an extension of 1.8 centimeters. The results obtained herein are discussed in terms of this specification. This report is divided into three phases. Phases 1 and 2 discuss test results obtained with the new diaphragms shortly after their original purchase in Phase 3 discusses test results obtained during the past month preparatory to a changeover to the new-type diaphragms.

7 Fourdrinier Kraft Board nstitute, nc. Page 4 Project Progress Report 155 MATERALS n Part 1 of this report, samples of pouncing paper (Cal 15) and combined board (CS 113) in current use in the laboratory calibration program were used. n addition, samples of 42-lb. kraft liner and 2-lb. test combined board were used for "waste" tests in this phase. n Part 2 of this report, samples of 42-lb. kraft liner submitted in connection with the liner baseline study during November and December of 1958 were evaluated. Part 3 employed the standard samples of pouncing paper and combined board samples as well as the 42-lb. liner samples submitted during February of this year. All materials were preconditioned for 24 hours at less than 35% R.H. and 73 F. and conditioned for at least 48 hours at 5 ± 2% R.H. and F. prior to test.

8 Fourdrinier Kraft Board nstitute, nc. Page 5 Project Progress Report 155 PROCEDURE PART 1. TESTS ON FRST LOT OF DAPHRAGMS Six diaphragms were evaluated as follows: 1. nstall the diaphragm and adjust it to give to 45 p.s.i. pressure at 1.8-cm. distention. Test standard samples used in calibration program. Note diaphragm pressures at start and end of tests. 2. Perform 1 tests on a "waste" sample of 2-lb. series combined board recording only the diaphragm pressure after every 1 tests. 3. Test standard samples as in (1) above. 4. After allowing the diaphragm to recover overnight, repeat steps 1, 2 and 3. All tests were performed on one tester (DE 35) by one operator. Twenty tests were performed on the pouncing paper sample and forty tests were performed on the combined board sample at each time. PART 2. COMPARSON OF BURSTNG-STRENGTH READNGS ON 42-LB. KRAFT SAMPLES USNG OLD- AND NEW-TYPE DAPHRAGMS DURNG OCTOBER AND NOVEMBER OF 1958 n this phase of the study, one of the nstitute's burstingstrength testers was removed from regular service and a new-style diaphragm was installed. After adjusting the initial diaphragm pressure to to 45 p.s.i. g., the twenty samples of 42-1b. kraft liner selected during November were evaluated. Duplicate tests were performed using the other nstitute tester with the old-type diaphragm. Twenty-four tests (12 up and 12 down) were made on each sample on each machine. The same procedure was followed in evaluating the December samples.

9 Fourdrinier Kraft Board nstitute, nc. Page 6 Project Progress Report 155 PART 3. CONTROL CHART EVALUATON OF "NEW"-STYLE DAPHRAGMS AND COMPARATVE TESTS ON KRAFT LNER DURNG FEBRUARY, 196 Preparatory to a changeover to the "new"-style diaphragms, each of the two bursting-strength testers in turn were removed from regular service for about two weeks. During the initial two-week period, the samples of pouncing paper and combined board currently used in the instrument calibration program were evaluated at daily intervals. n a few cases, longer intervals occurred because of the demands of other work. Twenty pouncing paper and forty combined board tests were performed on each day. After completion of the above, one of the testers ( DE 35)-- with new-style diaphragms--was used to evaluate each Project liner sample received during February, 196. These results were compared to the regular tests on these materials using tester DE 15 with "old"- style diaphragms. For each sample, 24 tests were performed with each type of diaphragm. The mill code used in this section of the report corresponds with that used in Progress Report 154, Project

10 Fourdrinier Kraft Board nstitute, nc. Page 7 Project Progress Report 155. DSCUSSON OF RESULTS PART 1. TESTS ON FRST LOT CF DAPHRAGMS As noted in the preceding section, bursting-strength tests were performed on each diaphragm using the standard samples of pouncing paper and combined board used in the nstitute calibration program. The control chart averages and limits for each material (based on previous testing experience) are summarized below: Pouncing Paper Combined Board Average, p.s.i. g sigma limits, p.s.i. g. upper lower sigma limits, p.s.i. g. upper lower With the above in mind, the results obtained with the new-type diaphragms are summarized in Table. Referring to the table, it may be observed that only two of the averages on pouncing paper fell outside of the 2-sigma limits. With regard to the combined board results, it may be noted that six of the values fell outside the 3-sigma control limits. These values were all on the low side and were associated with three of the diaphragms-- EP 7, 9 and 11. on the low side. n addition two averages fell outside of 2-sigma limits Referring to the over-all averages for each diaphragm. it may be observed that Diaphragms EP 7, 9 and 11 gave results about 8 to 9 p.s.i. lower than the control chart average. The remaining three dia-... phragms gave results which were in good agreement with the control chart

11 o 1 Av. o. '1 ct c H Av. S? o ~ X Oq ~ X 1. S 44 X CO %n m TABLE EVALUATON OF NEW-TYPE BURSTNG-STRENGTH DAPHRAGMS Diaphragm Number Cycle of Events EP 7 EP 9 EP 11 EP 14 EP 18 1st Day 1. nitial standard sample tests a. Pouncing paper Av. test, p.s.i. g a Pressure, start end b. Combined board Av. test, p.s.i. g b a Pressure, start end "Waste" tests on combined board Pressure, start 42 42, end Standard sample tests a. Pouncing paper Av. test, p.s.i. g a 15.8 Pressure, start end b. Combined board Av. test, p.s.i. g b b Pressure, start end EP

12 Av o -. a- CL 3 CD 1 m Ma c+ c1 C+- (D *o C+ NO "- \D TABLE (Continued) EVALUATON OF NEW-TYPE BURSTNG-STRENGTH DAPHRAGMS Cycle of Events EP 7 Diaphragm Number EP 9 EP 11 EP 14 EP 18 EP 24 2nd Day 4. Standard sample tests a. Pouncing paper Av. test, p.s.i. g. Pressure, start end b. Combined board Av. test, p.s.i. g. Pressure, start end b "Waste" tests on combined board Pressure, start end Standard sample tests a. Pouncing "-^sr Av. test, p.s.i. g. Pressure, start end b. Combined board Av. test, p.s.i. g. Pressure, start end b b a r7. Over-all average, p;s.i. g. Pouncing paper Combined board a Beyond 2-sigma limits for this sample. b Beyond 3-sigma limits for this sample.

13 Fourdrinier Kraft Board nstitute, nc. Page 1 Project Progress Report 155 average though slightly lower. The over-all average for all combined board tests was This average is below the control chart average by about 2%. Referring to the diaphragm pressure measurements in Table, it may be observed that pressures fell below the p.s.i. g. lower limit quite quickly. Using this criterion, three diaphragms (EP 7, EP 9, and EP 24) gave pressure readings below p.s.i. g. on the first day after less than 22 tests. The remaining three diaphragms failed to meet the pressure criterion on the second day after less than 4 tests. Therefore, a relatively short "life" appears to be associated with the new-style diaphragms, if present diaphragm pressure specifications are maintained. PART 2. COMPARATVE TESTS ON 42-LB. KRAFT LNER USNG OLD-AND NEW- TYPE DAPHRAGMS DURNG NOVEMBER AND DECEMBER, 1958 As mentioned previously, twenty samples of kraft liner submitted in November of 1958 were evaluated using the old- and new-type diaphragms. A similar procedure was followed in December of The results obtained are summarized in Tables and. t should be kept in mind that the comparisons include not only the effects of any differences due to diaphragms but also any effects due to differences in testers. Referring to the composite averages for the 2 samples, it may be noted that the results for the two types of diaphragms on the two testers were in reasonable agreement. n addition, most of the individual differences between tests appear to be within the range of normal variability. For eaample, in November only two samples exhibited large differences..these were file number where the-difference-was +8 p.s.i-. g. -and- - file number 1849 where the difference was -6 p.s.i. g.

14 Difference, p.s.i. g. q CD. - CD -1 *1 oq CD co w CL H CA. CD C, -t '1' -n '4' -" TABLE BURSTNG-STRENGTH RESULTS ON 42-LB. KRAFT LNER DURNG NOVEMBER OF 1958 File No. Bursting "New" Diaphragm (Tester DE 15) Maximum Minimum Average Strength, p.s.i. g. "Old" Diaphragm (Tester DE 35) Maximum Minimum Average o Average

15 Difference, p.s.i. g a. '1 ( D '1 C+-'* o '1 C,- tm '-1 CD ' *1 ~1 Cs. Cs. r Ct CD C) '1' c*oq m1 TABLE BURSTNG-STRENGTH RESULTS ON 42-LB. KRAFT LNER DURNG DECEMBER OF 1958 File No. Bursting "Old" Diaphragm (Tester DE 35) Maximum Minimum Average Strength, p.s.i. g. "New" Diaphragm (Tester DE i5) Maximum Minimum Average ill Average

16 Fourdrinier Kraft Board nstitute, nc. Page 13 Project Progress Report 155 While it was not considered necessary to test each difference statistically, the recent analyses of linerboard variability performed in connection with Project provided data for estimating at what level the differences might become significant ( ). n that study, the average standard error of 24 bursting-strength readings on 42-lb. kraft liner was found to be 2.25 p.s.i. g. By definition, the statistic "t" equals the difference between two averages divided by the standard error of the difference. that is, t = X1-2 SED where SET = J(SE1)2 + (SE_2) 2 and SE 1 and SE2 are the standard errors of the two samples f the assumption is made that the standard errors of the two samples are equal, then SE D = SE. On this basis Xi X X1 - X (2.25) at the 5 level, t = 2.14 for 46 degrees of freedom; therefore, X1 - X = (2.14) (3.118) = 6.3 The analysis suggests, therefore, that differences must equal or exceed 6.3 if they are to be significant at the 5 level. On this basis, in November only the difference of +8 appears to be definitely significant, while the other large difference of -6 borders on significance. The remaining differences appear to be well within the range that would be encompassed by the normal variability. n December only the difference of +7 for file number appears to be of possible significance.

17 Fourdrinier Kraft Board nstitute, nc. Page 14 Project Progress Report 155 PART 3. CONTROL CHART EVALUATON OF "NEW"-STYLE DAPHRAGMS AND COMPARATVE TESTS ON KRAFT LNER DURNG FEBRUARY, 196 Before initiating use of the new-style bursting-strength diaphragms on Project samples, two steps were taken. First, it was decided to begin obtaining control chart data on each machine--the data to be used in the nstitute's calibration program. Second, after this program was well underway, it was planned to evaluate all baseline liner samples received during February using both new- and old-style diaphragms. The control chart results obtained at this time are summarized in Table V. With regard to tester DE 15, the results indicate that the new-style diaphragms may give results slightly lower than those obtained with the old-style diaphragms--both in terms of test average and in variability. The percentage reduction of 4.3% for the combined board sample is, perhaps, greater than may be desired. On the other hand, both old and new diaphragms gave closely equivalent results on tester DE 35 in terms of test level and variability for both materials on the basis of the over-all averages. One result of the above is that the two testers are in reasonable agreement on pouncing paper but differ by more than 4% on combined board. Finally, it may be remarked that two diaphragm changes were required during the period for DE 15--despite the fact that no other testing was being performed on the machine. For tester DE 35, four diaphragm changes were required up to February 29. n part, the greater number of changes required in DE 35 reflects the fact that liner baseline tests were also being performed during this period. t is believed, however, that diaphragm changes will be required far more frequently with the "hnew"-style diaphragms than with the old-style diaphragms--if present

18 o cd..- o 1 OZco wcd t 'o Cr Cr oh'ā UQ 1 ' W o Vj'-fl CD TABLE V CONTROL CHART RESULTS OBTANED WTH "NEW"-STYLE DAPHRACMS Date ; a a Tester DE 15 Tester DE 35 Pouncin r Paper Standard Combined Board Standard Pouncing: Paper Standard Combined Board Standard Avaverage, Deviation, Average, Deviation, Average, Deviation, Average, Deviation, p.s.i. g. p.s.i. g. p.s.i. g. p.s.i. g. Date p.s.i. g. p.s.i. g. p.s.i. g. p.s.i. g a a a a Composite Average Averages (old-type diaphragms) 14? Difference, %b a Diaphragm changed. bbased on averages for old-type diaphragms. j

19 Fourdrinier Kraft Board nstitute, nc. Page 16 Project Progress Report 155 diaphragm specifications of to 45 p.s.i. g. at 1.8 cm. extension are to be maintained. This may be an important factor--particularly for control work where large number of tests must be performed each day because of the costs involved in changing diaphragms and rechecking testers. The bursting-strength results on the February liner baseline samples are summarized by mill in Table V and the data for individual samples are tabulated in Appendix A. As for the similar comparisons in Part 2, the differences will reflect both machine and diaphragm effects since the old- and new-type diaphragms were employed in different testers, i.e., new-type diaphragms in tester DE 35 and old-type diaphragms in tester DE 15. With this in mind, the results in Table V indicate that, in general, higher test results were obtained using tester DE 35 with the new-style diaphragms. The differences in p.s.i. g. for individual mills ranged from for Mill D to +6 for Mill P. On an over-all basis, the new-style diaphragms gave results about 3 p.s.i. g. higher than the old-type diaphragms. Parts 1 and 2 of this report and the control chart results discussed previously seemed to indicate that the new-type diaphragms gave results about equal to or perhaps slightly less than those obtained with the old-type diaphragms. The over-all increase of 3 p.s.i. g., noted in Table V, was, therefore, somewhat surprising and, if borne out in future testing, would produce an appreciable change in the apparent burstingstrength level maintained by the industry. With regard to the individual sample differences, Table V shows a frequency distribution of the differences. The greater number of the differences are probably not statistically different. However, if the

20 Fourdrinier Kraft Board nstitute, nc. Project Page 17 Progress Report 155 TABLE V COMPARSON OF BURSTNG-STRENGTH RESULTS USNG OLD- AND NEW-TYPE DAPHRAGMS ON 42-LB. KRAFT LNER SAMPLES FOR FEBRUARY, 196 Mill Bursting Strength, p.s.i. g. Old-Type New-Type No. of Diaphragms Diaphragms Samples (Tester DE 15) (Tester DE 35) Difference, p.s.i. g. A B C D E No samples submitted F G H J K L M N P Q S T No samples submitted Average

21 Fourdrinier Kraft Board nstitute, nc. Project Page 18 Progress Report 155 TABLE V DSTRBUTON OF NDVDUAL SAMPLE DFFERENCES Difference, p.s.i. g. +14 to to to to to to to 3.9 to to to to -4.1 Number of Samples Per Cent Total

22 Fourdrinier Kraft Board nstitute, nc. Page 19 Project Progress Report 155 criterion developed in Part 2 is applied (differences significant at 5% level if they exceed 6.3), 19 of the 115 sample differences or about 17% would probably be statistically significant.

23 Fourdrinier Kraft Board nstitute, nc. Project Page 2 Progress Report 155 LTERATURE CTED 1. Variability of kraft liner. Project , Progress Report One, November 7, 1958.

24 +f' ON N -4 ' N 4) ton -., 4) j 4) ~ -4 4).- a4 a) d) 4) em\ emn UP i \ -em T em a, 4 -M n co,- r4p.i -4.? - F-1 *r4.v- C co F -,4-4 m m~r " m t - ~~~~o j H 4-) 4-) -H ( -4 m 2 K4 -ct.cc co -4 -) m V co : r~ a, 43 ( H a) ) z. Ḥ r- '' C V'\\ V' r-wwwi cicion o W~Ni N V' fl)n em% mn em ~ln fl ONC tl mci 'n'n OOO2 O O ~oo U mn4o--t t'-. NC~ f 22 ONOD 'cow tflr \ a OD em N~ C) ' H.4- Nem4C" t o em'o N MU) N fno Nv -4\N ON 'rl 4Q em 'ni eme -4\co \ c m 43 b4- td em NN Cl 4-* 'flt'2 %- fn N~ N\ N~ (\ N N~ \ NT o QCjc \NJ\ 4-- flnnn.4 '4 N (MNC *r (i.. - N.4 em4 wm' o CN o Cy oo P TCK 4a co Li c 4t ZS~pf NC i '(7-4 N~ \ Oj ~ o -4-4 ~?- t a

25 en O N N 'nf ) ) C) ho P. a: a) () U) (D-i ) ~i- 'HA 4 A.~fMbO-4Q\ 4V N~ ~.-44u' (r~ ho t1 1-4 dtc 4 ' fl' t N ox,-4 41 ~ "'- OD o N.4 H C r. - t~ 4, CO Pi H H- to' N en a '\ n".4 \ V\) Nflf ccl (Nn ') N~ Nl H- Hl %'%C"%\'fl'fle a, N O O (7 No - N N nnmmhnnn \ en m Cx. H- ih enen W N\ ecoe TN OQJ +3 -H (4-4-4 la : E- Z- zo H El a) H~ 3 ~$\ 4-' C a'l a'.to'e o'z no 4~C nen tc\ N" N* (fcn O N \ N, N~ - onong 'fl.-co enn-:t o\'.o\o o CD C\ O WM C\ Co UW% \fln C a 'r, U (\-- t -4- \ C4 ' en en NQ flnw -4 NN'- N \ 4~ co-4-4 Eri. 54 ).9-i Er 4 "weo ~ '-* cn ) hko cu S. 4) -'c 'fl'f c \.. N (' flfl :4 ) 4 4 "'t f-,flnfln en w e O ) :4

26 ~ N 'low NtfllflN Wo '. NONON NCNCnn nfl% M -4 nn 'Dn fl4 ('- NON ---- co t 4-- o C 4(N N' - C 4N CNc fic C\O NN C C)O\ -toc's C" l~c" 1 N 4n\n ~~~~~~~~~~~~~~~~~~~~~~~~ -4 \ ' -4\ ', Cd 54 fl- 4 a) 4 '-4 Cfl- N-tj--O n-4) N~ Cd C.1 C) H 4-3 ~1 -H~ C) H C p.4-4 C). N 'C2'l "r\o\ no c-%\ 'cc) Cy\ 'no'c ' ' co \ONOOcJOONN N CG'. o~ c'\ a\'ciw (N ON.4- CO> c-. < C CYn 'flfl' N fl,'fla n -4N -- 4-V,4C~ -- 4t N~ (H z HZ (DC ccc OD'C a' ci\ i \ \f ~ 'D lo -'r~ w ( flflflfn - CCiN f~ CN \ 4C-t ~.(--- NNOC ~ Hi ' 54 Cd 4-, 44 ( HOMOnN NOD o \ co Mr N N N fl-\cr'lcmn '1 Mr 5 a). S., ' 54 C; z (±4 Cd a) -cc cicici OD OD OD i i co C) CD 'flq C'-ODW'N Cr co C) '.'Q G P O R' r M -C

27 ii) a. ) a) '4.r 1) Q- '4N CO '4 H n 4 r- ci tn N N.. C4 H 2 a. 'c4 F < MO. Cd. 'fl1-h C ~ bi) (; bdzd.h :i ~O'ONO'o Cm C"% co-4 m- i O\ N zl -4-4o N- N~ -tc-u- ~N~ N C ( n n, i- -H fl-, xj CO co O'\ c ) c ' \ N N H ' $4 Cd 92 C-i Cd f 2 5 )) ; H - CON c o 7\a C\N N (flfl N cc a, 4J N 4 MCflN~f ODcflaiOD O NCc (flflcn CN N\ fl' \ (f emo OD G\CO O c coc 444 ON'-t(7 em- cdc% - N' N ~l' ' C.4 4, H~ ' d 14 -~O N \fl4!? 2?!9!Y co coco c ) -C ' ON) r4c ul% ho c4..

28 -44t '-4- n- + bd-p. C h 42) ( a) -. U) S- -H '. p- ' NNC, +!P i+ + -\4 H.C co a Q V,\mm4w " N - (\J... +p a, J H ri 4-, ( -4~ () F-. < U) C) mi,a. bdz. 4)-. C) H" 9-41 ono"nktu l Q -4 C N OmC\ ON Y\N 'o CY,.*' N N WMN -4 fin Cy ) c CON(\ ON C'.- Nl N N MiN fin nn, O D n --' " (" C>..., " - NP -4l coa O H ON r 4, C- co 2 : CH bd. at.a En ; z r.-. o N Cy "O\ N yg ~KŽ~OOC~2# ~~~~~- -4MoC~C\c NC -4 2) N Nl D to Ct 1. C) c \- G\ 'H 'HO\ C) -' ~O f%(,o\n- - ' -4 'co -4N N ~~~%. 'N' co co.. 44 \ O\ -4 A. 'N NC NO. No c' N" C-4-' C) t. CD -~.T~ 4 SZ - *~ -..

29 -4 en ' en C-4-4 t4 '.-4 a) a5 a). d) a) $5 -H 4). '-4en4 +++ ' (r + '-4' fl-2- N a). 2 H -4 ). N 'ClN ' ~ o~ r co 'fl4 C NO o-'o, ~ ~ ~ -4 C2 z :> o 'nlo 4ONON N, '\ OD N~ C, C3 \ co \ 34- CO ' co N C4 2 co'd -4r, CC P Y H 4) 4-, 14 4, 1 ~ 54 -c4 r. H F. H- CZ 4¾C 4Q~ 4)w P ri4 Hio 'O\3 co N)C nnc' -2-C\ en "N -4 \ \ '4 fc~e C,3 '-t1~\' ' en 2-1 ) -4j C-\ en Hi, N CenMNc C\2 C"\ '\ N~ N.-4 OD 3 OD 3 co NQ N~ N N (N O b4'\ ' 3 N C4 to'-,f~ H i ' - tf ' U'noC' en -2- en en, en% en C\ -4-4 r t--t CO CO- 4. ~ ~ o o N- N ~ en 'r'. en\ N.-t entn flj N ~ 4' (-H $ 54el a).14- x.. N3 co C N~ -4--\ ' 33333T '--C co4 co oo a) -4C) a,'o ' ON \ ' ---- ~~~~--t to -~c N~ en -4 ("A CO atr ' n ' n ' O \ m ' -4 -., cz -'

30 U~~floe JG8tRUO ~ ~ U ~uo 4 ;aode-d ssaa~o.a a~ed r-st u1 Gfl;;;Tsu] pzuoai 4jeay fttlou;.rp-n

31

32 !..j k. t - 11 t e f 1, i 4. 1., '. 7 ).i i.,: 11 ;,, -l :,., : f 1,..!,;. l -..i t. ; 1!; j. 4. <-'-, 4~~~~~~~~, :..... ".. ; 1!.:. "'Y': ) i.!, j!;, i. i 1; t;.: j.., 1,,, il~ ia, e. -i *t1 i t i i 5 h i,. :, : t :,,7,, :,, i -, 3 i11 i, i.,. 1;,1, i V. t t. i.! i 1 ;i ; :,i ( : i,., i -4,, i ; 2:,,!., li i, 1 i. i i: 11i. i i. z,, 11, 2A 4 '. 4 ~ 4- i