172 BULLETIN OF THE UNITED STATES FISH COMMISSION. 



larger, but being lighter, are in part deposited in a thin layer on the top of the mass 

 of larger organisms. We may consider separately the errors which are introduced into 

 the volumetric method from the three sources above mentioned. 



(a) Errors due to rloi/yiny of the net. This depends principally upon the area of 

 the filtering surface of the net as compared to the volume of plankton present in the 

 water. If the net surface is large and the volume of plankton in the water liltered 

 small, there is but little clogging. The net employed by Kotbid was 25 cm. in diam- 

 eter at the base and 40 cm. on one side. The plankton appears to have been unusually 

 abundant (Kofoid gives no data) and the conditions otherwise unsuited to the use of 

 any sort of net. The net employed by Reighard and Ward in the work above referred 

 to had a diameter of 00 cm. and a slant height of 100 cm. Its filtering surface \vas 

 thus about six times that of the net used by Kofoid, while the plankton in the water 

 in which it was used was very little. In the work done by Reighard not more than 

 4.5 c.c. of plankton was taken in the net at one time and in the work of Ward not 

 more than 11.9 c.c. In a majority of the hauls not more than a fraction of these 

 volumes was taken. The net used by Hensen was much larger (Hensen, loc. eit., p. l>), 

 while that used by Apstein was about the size of Kofoid's net, but it was probably 

 used under more favorable conditions. Clogging, then, does not seem to me to be an 

 important factor with nets of the size used by Hcnseu, Reighard. and Ward. It 

 becomes important only in case a small net, such as Kofoid's, is used under unsuitable 

 conditions. Some measure of its extent is desirable. 



(b) Error due to shrinkage. This error is largely if not wholly eliminated by 

 previous thorough shrinking of the net. The cloth used by Reighard and Ward was 

 several times dampened and ironed before it was made up into the net and was thus 

 presumably thoroughly shrunken. The net was also many times wet and dried before 

 it was used for quantitative work. As may be seen from the table on page r>7 of 

 Reighard's report, the cloth of the net used by him and later by Ward differed but 

 little after a summer's use from new cloth which had been once wetted and then dried; 

 the cloth in the two cases being measured under as nearly as possible the same condi- 

 tions. Whether the nets of other workers were similarly shrunken before use does 

 not ap{>ear. I have not encountered any such enormous shrinkage as that recorded 

 by Kofoid, in which the average size of net openings was reduced from .000024 to 

 .00001 sq. cm. Everything here depends on a uniform method of measuring the cloth. 



(c) Errors due to iimncdhiliti/ of the cloth. A large number of the smaller plankton 

 organisms escape through the pores of the cloth. According to Kofoid "the. silk 

 net retains from to A of the total solid contents of the water." "The amount 

 escaping through the silk bears no constant relation to the amount retained." These 

 statements are certainly very startling, but one. must reserve final judgment concerning 

 them until the conditions of the experiments upon which they rest are made known. 

 This degree of leakage through the net may be, due to the peculiar constitution of the 

 plankton examined. The extent to which this source of error vitiates previous work 

 can only be determined by tests of the nets used by previous workers in comparison 

 with other methods and in the waters in which the nets were used. In volumetric 

 determinations most of the smaller plankton organisms are packed between the larger 

 organisms in such a way as not to affect the total volume of plankton in the measuring 

 tube. Some of them, however, remain in suspension longer than the larger and heavier 

 organisms, and when they settle lie at the top of the whole mass measured, and so 

 increase its volume. 



