96 FRESHWATER ANIMALS 



of something of the order of a million individuals.) The sam- 

 pling was done so that all depths of the lake were equally repre- 

 sented at all dates and so that each series of samples was quanti- 

 tatively equivalent. In equivalent columns of water there were 

 three specimens of retrocuwa on April 1. By the end of that 

 month there were over three thousand individuals, and the 

 number remained between three and five thousand until the mid- 

 dle of June. By the end of June this had dropped to about one 

 thousand, and on July 21 and August 10 there were less than 

 two hundred retrocuwa in the standard column of water. These 

 population densities are of interest in relation to the phenotypic 

 variation. 



The appearance of the individuals that occurred at the dates 

 of sampling is indicated in Figs. 3 and 4. The proportions of all 

 undistorted specimens in a sample (up to a maximum of about 

 one hundred) were measured, and the figures drawn in each 

 horizontal row depict the average proportions (of head vs. cara- 

 pace) for each size grouping that occurred. By studying the 

 distribution of size and proportions in this temporal series of 

 population samples, it was possible to estimate whether or not 

 the majority of the newborn in the population sample represented 

 a new generation or merely successive broods of the same gen- 

 eration. For example, on April 15 the newborn individuals were 

 almost certainly the offspring of the over-wintering females ( seen 

 as the largest females on both April 1 and April 15 ) . The female 

 with a carapace length of 865 microns (CL = 865) represents 

 the most numerous size group. These are the individuals which 

 were the first-born of the year, and which were approaching 

 maturity. However, the old, large over-wintering females were 

 the only ones producing eggs, and these in large numbers. It 

 was not until the end of April that the second generation pre- 

 dominated among the newborn. Although such an analysis lie- 

 comes less clear-cut later in the season, it is reasonable to esti- 

 mate that there were seven generations during the four-and-a- 

 half-month period of observation. On this basis one can make the 

 even rougher estimate that there are probably about ten or 

 twelve generations of Daphnia retrocuwa per year in tin's lake. 



