years; it was 17 per liter in 1929 and l5 per liter in 1930. 



Rare forms. The distribution of Diaphanosoma was similar 

 in the three years studied. It was somewhat more abundant in 1930 than 

 in the other two years. It is essentially a summer form, with a 

 maximum in July or August, 



The remaining groups of Crustacea were so rare in all years 

 that comparison of the counts has little value. Limnocalanus was 

 apparently more abundant and had a longer season in 1930 than in 

 1929, but was rare in both years. It is distinctly a cold-water 

 form and is absent during the hottest period of the year. Epischura 

 was present during all of the sampling season in 1929, but was not 

 taken in trap samples after July, 1930. In 192 8 it was present only 

 occasionally. It rarely occurred at a station to the number of 1 

 per liter. Bosmina was about equally rare in all years. It had two 

 periods of development, one in early summer, the other in late 

 summer. Because of its small size and rarity, Bosmina probably has 

 little importance in the plankton. Limnocalanus, Leptodora, and 

 Epischura are rare also, but their large size gives them importance 

 as food organisms. 



Adequacy of the sampling program 



The question of the adequacy of the sampling program may 

 well be taken up at this point. Only the data for I929 and 1930 

 will be used for purposes of illustration. It has been shown that the 

 zooplankton is not uniformly distributed in the Island Section. In 

 order to avoid inaccuracies arising from this condition, an attempt 

 was made to sample several stations in each two-week period. In 1929, 

 late July was represented by only three stations, early August by 

 four, early October by one, and late October by two. The other 

 periods were represented by seven or more stations. In 1930, the 

 two periods in April were represented by two stations each, and the 

 two periods in July by three stations each. The other periods were 

 represented by five or six stations. 



If this program of sampling were inadequate, we should 

 expect the points on the curves for seasonal distribution to fluctuate 

 up and down with no evidence of seasonal trends. Reference to Fig. 21 

 will show that each of the eight curves was low in spring and fall, 

 and high at some time in the summer. Moreover, with minor exceptions, 

 the fluctuations in abundance lacjj the appearance of being fortuitous. 

 That is, in general, successive points show a progressive increase, 

 decrease, or maintenance of abundance over a considerable period of 

 time. The principal exception to this statement is seen in the curve 



226 



J 



