J. L. BROOKS 93 



the genetic diversity during the periods when the population is 

 reproducing by diploid eggs. 



As the determinants of this phenotypic variation are not ge- 

 netic, they must be environmental. The simplest hypothesis is 

 that the size of the crest or helmet is dependent on the environ- 

 mental conditions during the growth of the individual concerned. 

 In order to study this relationship it is desirable to have a popula- 

 tion showing an extreme range of phenotypic variation in a 

 natural situation in which the population's environment is as 

 homogeneous at any one time as is possible. 



A fortunate combination of these desiderata occurred in Ban- 

 tam Lake, Connecticut. This is a large, shallow lake with three 

 basins, the largest and deepest of which is about a mile in diam- 

 eter and seven meters deep. In such a body of water, the water 

 is likely to be nearly the same temperature from top to bottom 

 because the winds are able to mix the water throughout most of 

 the season when there is no ice cover. As there was evidence to 

 indicate that temperature plays a major role as a determinant of 

 the phenotype, such a lake is to be preferred over one in which 

 the warmer upper strata are underlain in midsummer by strata 

 of cool or cold water. Although the behavior of most of the spe- 

 cies that can develop helmets is such that they remain, day and 

 night, in the upper strata, the possibility that a portion of any 

 population might have penetrated into the cold, lower waters 

 would complicate the analysis of the environmental effects. In 

 the year with which we are concerned the principal species of 

 Daplinia was D. retrocurva Forbes. This species exhibits prob- 

 ably the most extreme variation of any North American species 

 and one of the most striking in the genus and, indeed, in the 

 Cladocera. 



The population of D. retrocurva in Bantam Lake during 1945 

 was sampled at approximately two-week intervals from April 1, 

 one week after the ice melted, until August 10. By the middle of 

 August the population had so dwindled in size that sampling 

 yielded only a few specimens. (However, it must be remembered 

 that even a population so sparse as to yield but a few specimens 

 to this standard sampling procedure could amount to a total size 



