554 
FILTER-PAPER CATCHES, ILLINOIS RIVER—Continued. 
Z. wae ce tc meee || 
Le 2 ae we © Oa 
28 2 £2) 6 Bee) bee!) Se ee el eee 
ne a Soy = ae Bow Ss (ou, Sus 
“a = 
33 5 ees g 8 |eee | 68 | ga | Se 
< = a Ho, c= 
1899 
22,821 d | 3, I. | 1000 04 40 82.00 | 32.80 7.20 | 0.22 
22,826 d | Io, I. | 1000 .025 25 8€.00 | 21.40 3.60] 0.15 
22,828 d | 17, I, | 1000 05 50 95-50 | 47-75 25 | 0.50 
22,831 d | 24, I. | 1000 +045 45 99.00 | 44.55 -45 | 0.03 
22.833 d | 31, I. 1000 .O4 40 99.00 | 39.60 .40 | 0.01 
225838 (Gazelle 1000 .05 iXe) 98.00 | 49.00 1.00} O.I1 
22,841 d | 14, II. | 1000 05 50 |. 90.00 | 45.00 5.00] 1.15 
22,845 d }21, II. | 1000 05 50 60.00 | 30.00 | 20.00; 1.92 
22,847 d | 28, II. 500 41 820 99.99 | 819.92 08 | 0.07 
22,850 d | 7, III. 500 .20 400 99.90 | 399.60 40 | 0.54 
22,853 d | 14, III 1000 16 160 99-90 | 159.84 16 | 0.35 
22,856 d | 21, III 1000 28 230 99.80 | 229.58 42 | 0.21 
22,858 d | 28, III 1000 04 40 99.80 | 39.92 08 | 0.01 
AV Cla MEA atrccty tie ricci t ieee eines 3.16 | O.41 
undetermined extent invalidated by the errors above noted, I 
wish to call attention to the fact that they may still serve to 
indicate in some degree the extent of the leakage and its sea- 
sonal distribution—conclusions which are in some measure 
corroborated by the results of enumeration. The ratios of the 
volumetric determinations of the plankton by the silk and fil- 
ter-paper methods in August-December, 1896, in 1897, in 1898, 
and in January-March, 1899, as shown by the averages, are 
respectively 1 to 4.1, 1 to 3.6, 1 to 2.6, and 1 to 7.7, or, averaging 
all collections, 1 to 3.3. If these figures approach the actual loss 
by leakage it becomes a matter of some volumetric importance. 
An examination of the table reveals the fact that in a third 
of the cases the estimated plankton in the filter catch is exceed- 
ed by that of the silk net. It will be seen that most of these 
cases occur in instances of small plankton, where the tendency, 
above noted, to overestimate the silt is most effective in caus- 
ing this apparent deficiency. In all cases the total filter catch 
greatly exceeds the total silk catch per m.’ (cf. Table III. and 
the one under discussion ). 
In general the preponderance of the filter catches is greatest 
in the warm season of May—September, the growing period of 
