390 Birge — The Crustacea of the Plankton. 



this level fell off in number even more rapidly than they had in- 

 creased — so rapidly, indeed, that no effect was produced by 

 these broods upon the population of the water below 3 m., ex- 

 cept perhaps to check in some degree the rate of decrease 

 toward the winter minimum. There was also a small rise in 

 December in the 0-3 m. level, caused by the increase of D. 

 pulicaria. 



It would seem from these facts that there is a maximum pop- 

 ulation per cubic meter beyond which the Crustacea are unable 

 to multiply and which differs in different seasons. It is difficult 

 to see what it is that sets a limit to this population in the 

 autumn. At this time the food is in enormous abundance as 

 compared with the number of the Crustacea, and it would be 

 expected that the numbers in all levels of the lake would in- 

 crease together. I am quite unable to give a reason for their 

 failure to do so, but the fact recurred exactly in all three years 

 of my observations, making allowance for the peculiar condi- 

 tions in the early summer of 1894. 



Fig. 28 represents the average percentile vertical distribu- 

 tion of the Crustacea for Oct. 1-15, 1896, March 1-15, 1895, 

 August 1-15, 1896. The corresponding figures are given in 

 Table C, appendix. In the diagram each horizontal space rep- 

 resents 10 per cent, of the Crustacea and each vertical space, 3 

 m. On each 3 m. line is platted the percentage of Crustacea 

 found below it, and these points are connected by a line which 

 extends from 100 per cent, at the surface to at the bottom. 

 From the intersection of these curves with the vertical lines can 

 be seen approximately the percentage of the Crustacea above 

 and below the depth indicated at the intersection. If the dis- 

 tribution were uniform there would be 16.6 per cent, in each 

 vertical space and the percentile distribution would be marked 

 by a straight line running from corner to corner of the diagram. 

 The curve for October approximates very closely to this, the 

 percentage being larger in the surface stratum and somewhat 

 smaller below 12 m., but, in general, the line lies very closely 

 parallel to the diagonal. The distribution for March is almost 

 equally uniform, but here the bottom level has an excess, due 

 to Cyclops, and the 0-3 m. level is slightly below the average. 



