392 Birge — The Crustacea of the Plankton. 



Each species of Crustacea, also, has individual peculiarities of 

 distribution, which recur from year to year with surprising 

 similarity and which are independent of the absolute number 

 present. These peculiarities appear when the average of any 

 species is taken, although of course it is entirely possible that 

 the distribution should depart widely from this average at any 

 single observation. In general it may be said that the summer 

 distribution of the Crustacea follows very closely the figures 

 which are given in my former paper (Birge, '95), and that the 

 variations in the distribution which have been found during the 

 two years and a half succeeding the observations reported in 

 that paper, have been of the same type and in general of the , 

 same degree as those which were found during the single month 

 of our first study. It seems to me, therefore, unnecessary to 

 point out again these variations in detail for each species. 



In order to show the resemblances and differences in the per- 

 centile distribution of the Crustacea during the summer months, 

 when their numbers are great and the distribution is most 

 characteristic, I have averaged this distribution for the 

 summers of three years: 1894, 1895, 1896. I have included the 

 three standard representatives of the limnetic Crustacea which 

 are regularly present in full numbers during this time; Diap- 

 tomus, Cyclops, D. hyalina. The period included is from the mid- 

 dle of June to the middle of September, in 1895 and 1896; and 

 July and August of 1894. It will be remembered that no ob- 

 servations were taken in 1894 before July or during the first 

 part of September, but as the summer conditions were thor- 

 oughly established at the first of July of that year and contin- 

 ued until the first of October no noteworthy difference would 

 appear in the averages had it been possible to extend the period. 

 It will be seen from these averages that the distribution of Cy- 

 clops in the three years in question varies surprisingly little; 

 the percentile difference in the 0-3 m. level being less than 

 1.5. This close correspondence in distribution exists in spite 

 of the fact that the numbers of the genus were very different 

 in the three years. The same general agreement is seen in 

 the tables of semi-monthly distribution. Compare July, 1894 

 and 1896 in Table C, Appendix. 



