216 



Marsh — Limnetic Crustacea of Green Lake. 



August, 1893 



August, 1896 



September-October, 1894 



October, 1894, day 



October, 1894, night 



0-5 



48.30 

 39.27 

 52.01 



38.28 

 69.07 



5-10 



30.60 

 39.89 

 16.28 

 17.89 

 9.84 



10-15 



19.60 



15.46 



16.54 



8.23 



15-20 



15.82 

 6.01 



I do not think that the number of collections is large enough 

 to draw inferences final in character in regard to the vertical 

 distribution of Daphnella, especially since the total number in 

 any collection is small. It would appear, however, that the up- 

 per five meters are more densely populated in September and 

 October than in August and that the number is also greater in 

 the upper five meters in the night time than in the day time. 

 I do not feel like speaking in any dogmatic way in regard to 

 the interpretation of these facts, but I venture to suggest that 

 Daphnella is, in its vertical distribution, controlled rather by 

 light and darkness than by changes of temperature. If it were 

 very sensitive to changes of temperature the fact that it is 

 found in greater numbers near the surface at night than in the 

 day time, and also in greater numbers in September and Octo- 

 ber than in August would indicate a liking for cool water: but 

 if this liking were very pronounced, it would seem that it 

 would migrate deeper in August. If we suppose light to be 

 the controlling factor, we would explain the greater number 

 near the surface in September and October by the greater num- 

 ber of cloudy days in those months. Very likely the solution 

 of this problem is not so simple as my speculations would indi- 

 cate, and a satisfactory result can only be reached by a care- 

 fully conducted investigation in the laboratory of the behavior 

 of the animal under different conditions of light and tempera- 

 ture. It may be noticed that Apstein ('96, p. 79) states that 

 the time when the larger numbers are found at the surface, co- 

 incides with the time of total maximum numbers, a conclusion 

 quite the opposite of what my observations would indicate. It 

 does not appear, however, that his conclusions were based on 

 any large number of exact observations. 



