R. C. PUNNETT 
321 
-^^ The Numerical Proportions of the Sexes. 
In connection with what has gone before a brief digression on the proportional 
numbers of the sexes may not be out of place. Storm's ('80) experience was that 
the cfs of Spinax were rarer than the $ s, though he gives no numbers. Records 
of the numbers of each sex caught off Ask have been made at different times by 
Braus, Nordgaard, and myself, and are given in the subjoined table. 
</ ? 
121 98 
37 83 
47 65 
IS 2B 
218 269 
The collected figures agree with Storm's conclusion, the only case where the 
(/s were more numerous being Braus' expei'ience in 1897. Braus also collected at 
Drobak, near Christiania, where he found that the J's were very much scarcer than 
the $ s, being in fact only half as numerous. There are good grounds therefore 
for supposing that in the adult state the cfs are less numerous than the $s. The 
sexes are, however, produced in practically equal numbers, for out of a total 
of 308-f embryos were found 149 cfs and 159 $s. And this conclusion is 
strengthened by the fact that in the three different years, 1901-3, in which 
embryos were collected, the proportion of cfs to $s was respectively 35 : 39, 
48 : 46, and 66 : 74. Consequently from a consideration of the proportional 
numbers of the sexes before birth and during adult life we are led to conclude that 
post-natal mortality is greater among the cfs than among the $ s. This conclusion 
is of considerable importance, for, as will appear below, statistical treatment of 
the meristic variations points independently in the same direction. 
*Braus, July, 1897 
Punnett, July, 1901 
June-July, 1902 
Nordgaard, July, 1903 ., 
II. Vakiation and Correlation. 
Meristic Variability. 
The measure of the variability used is the standard deviation (a) which has 
been calculated separately for the adults and embryos, and for both sexes. The 
results are given in Table 2, and allow of a comparison between the sexes, for 
embryos as well as for adults. As there is some irregularity in this table a rough 
approximation to the variability has been made by taking the mean for the first 
* In his paper on Spinax, Professor Braus gave tlie proportion of cf s to ? s as II : 5 at Ask, and 5 : G 
at Drobak ('99, p. 421). In reply to a letter of mine Professor Braus explained how an error had crept 
into the MSS. on this point and very kindly sent me the actual figures which I have here made use of. 
+ Four of these j s do not figure in the Appendix Tables, as they were crippled and therefore useless 
from an anatomical point of view. 
Biometrika iii 41 
