from near the bottom to two meters. In computing the number of organisms 

 per liter of lake water, it v/as assumed that the net strained one half of 

 the column of water through which it was drawn. 



In 1929 and 1930 all quantitative samples were taken with -the 

 plankton trap described by Juday (I916) . The trap has a capacity of hS 

 liters. One hundred and five series of samples were taken in each year, 

 which makes a total of 210 series. The total number of trap samples was 

 971. 



It has been assumed, for practical purposes, that the trap takes 

 a perfect sample; that it captures all of the plankton organisms which 

 were present in the hS liters of water before lowering the trap into 

 position. Actually, of course, lowering the trap has a distributing in- 

 fluence, and the active plankters tend to move away from the center of 

 disturbance. The lowermost part of the trap is the net, and this is at 

 one side of the box-like part which will enclose the U5 liters. If the 

 organisms move away from the net in all directions, the population is 

 increased in the block of water which an instant later becomes enclosed, 

 so that the trap captures more organisms than it should. This supposed 

 action of the trap has not been tested experimentally, and no account 

 has been taken of it in computing the number of organisms. In a way 

 such action would be fortunate, for in all subsequent handling of the 

 sample there is a tendency to lose plankton. However, this advantage 

 would be offset by the fact that the active plankters would be increased 

 relative to the more passive ones. 



In taking a surface sample the trap was lowered just far enough 

 to submerge it completely, and since the height of the effective part 

 of the trap was 50 centimeters, the sample would be a sample of the 

 plankton in the upper half meter of water. In taking a sample at 



2 meters, the bottom of the trap was lowered to that depth; hence the 

 sample would represent the layer between 2 meters and 1,$ meters. 

 After the water had strained through the net of No. 12 bolting cloth, 

 the catch was washed into the plankton bucket at the bottom, and then 

 transferred to a three ounce bottle. 



Because of the lack or uniformity in vertical distribution of 

 the plankton, it vas necessary to take samples at several depths to 

 obtain an accurate average for a station. The distance between samples 

 in a vertical series was ordinarily 2 meters, but in many cases it was 



3 meters. Only two series in which the interval was i; meters have been 

 used in this report. The question arises: should samples have been 

 taken at more frequent intervals to avoid errors due to differences in 

 abundance at different levels? Table 63 was designed to facilitate a 



202 



