shci0.d te pointed out, hovrever, that the smooth progressiorx of 

 increasing average ovum diameters with later dates of capture in 

 Tatle 3 is most lilcely fortuitous. In exajnining nearly a thou- 

 sand females^, I found that many degrees of maturity vrere repre- 

 sented among the specimens taicen on a single da;^^, particularly 

 in midseason. The mean values listed in Table 3 a^e considered 

 to represent the average degree of egg development in females 

 entering the stream on the dates indicated. 



In migrants entering the stream in mid-April, 63. Ll percent 

 of the development of the ova had still to talce place ivhile the 

 female was in the stream (computed on the basis of the relative 

 volumetric proportions of individual ova vdth the assumption that 

 they are perfectly spherical) . In females that entered the stream 

 from :nid-June imtil the end of the run, the eggs were very nearly 

 fully developed, but not so far along as to have burst into the 

 hody cavity. In only a few of the late migrants, including the 

 last ones taken in July, were any eggs found in the coelom. 



In the frequency polygons presented in Figure 2$ a uniform 

 horizontal scale Tvas used for all six projections. Thus the 

 relative degree of matui'ity of the specimens examined is apparent 

 in the progressive shifting of the frequency distributions from 

 left to right. Of particular interest is the great scarcity of 

 partially developed eggs found in most of the samples represented. 



(c) Egg production 



Of the 70 specimens utilized to determine egg production, 

 actual numerical counts were made of the ova in 10 specimens. 

 These included the largest (21.1 inches, $36 ram.) and the siaall- 

 est (12.6 inches, 320 mm.) females in the series and ei^t of 

 intermediate size. Prior to malcing each of these counts, the 

 total volume and the total weight of the ovary ivere obtained. A 

 sample section was removed from the middle of the length of the 

 ovary and the volume and weight of the sample were likewise deter- 

 mined. Volicnes were secured by a system of displacement of water 

 into a cylinder bearing 0.2 cc. graduations and vrere read to the 

 nearest 0.1 cc. Weights were obtained to the nearest 0.01 gram on 

 a shemiical balance. Excess moisture was removed as consistently 

 as possible from all ovaries and sections before any determinations 

 were made. Eggs in the sample sections were counted first and the 

 total production "vms computed by dii-ect proportion for both the 

 volumetric and the gravimetric data. When the baiance of tiie ova 

 had been counted, the calcu].ated totals by both techniques could 

 be compajred for accuracy. This procedure was follovred for all ten 

 specimens. 



78 - 



