Introductory. 
275 
cea was found in the course of the year 1895, it seemed to me 
advisable to continue the observations through the season 
of 1896, in order to determine whether the course of develop¬ 
ment would be the same as in 1895. Until August, 1896, the 
number of the Crustacea in each catch was determined separately, 
and the average catch for each two-week period was computed. 
After that date the catches for each two-week period were 
mingled together, and the average number only was determined. 
Up to August, 1896, therefore, the average, maximum, and min¬ 
imum catches are given for each period, in the tables of the 
appendix, but after that date it is possible to state the aver¬ 
ages only. This “two-week average” is the main number 
used in this paper. 
The net employed was that described by me in my former 
paper, and the method of counting was substantially the same, 
except that a smaller fraction than one-sixth was often used to 
determine the large number of Crustacea from the upper levels 
of the lake—one-tenth to one-fifteenth being ordinarily employed, 
with a view to making the last figure of the resulting number 5 
or 0, in order to facilitate adding and multiplying in subse¬ 
quent operations. 
The multiplications to reduce the catch to the number per 
square meter of surface were performed by the aid of Crelle’s 
Tables. The products are stated in this paper in thousands and 
tenths, in order to avoid the constant use of ciphers in the last 
two places. The result would have been quite as accurately ex¬ 
pressed in most cases if the nearest thousand had been stated, 
but in case of the smaller numbers it was necessary to state the 
hundreds, and as the products were read off directly in all cases 
in hundreds, I concluded to leave them in the printed results, 
although, of course, understanding that no reliance is to be 
placed on the exactness of the enumeration in the last place of 
figures if the total is large. 
The total number of serial observations was 333 besides 97 
single catches, and as there were at least six collections in each 
series, and from three to eleven species of Crustacea to be deter¬ 
mined, the number of single observations is very large — over 
10,000. It has been my aim in preparing this paper to exhibit 
