TABLE X. 



295 



of specimens having 12 dorsal spines, 

 of specimens having 13 dorsal spines, 

 of specimens having 14 dorsal spines, 

 of specimens having 15 dorsal spines, 

 of specimens having 16 dorsal spines . 

 of specimens having 17 dorsal spines. 



First 

 Year. 



Second 

 Year. 



Third 

 Year. 



Per cent. 

 Per cent. 

 Per cent. 

 Per cent. 

 Per cent. 

 Per cent. 



1.69 

 38.98 

 50.00 



7.62 



0.84 

 8.47 

 49.14 

 35.16 

 5.50 

 0.42 



3.65 



36.17 



51.62 



8.13 



0.40 



The first and third year specimens resemble each other very closely in regard 

 to the scales in the lateral line and the dorsal spines. In these characters the 

 second year specimens show a decided difference. These have on an average two 

 more scales in the lateral line, and have 14 as the prevailing number of dorsal 

 spines instead of 15, the number in the first and third year specimens. 



Several explanations might be suggested to account for a part or all of these 

 differences. 



The explanation suggesting itself most readily is that an additional spine and 

 ray are added during the life of the individual. I have gone over all the specimens 

 carefully with this point in view, but find no evidence either of the splitting of a 

 ray or spine, or of the new growth of these, except at the anterior of the dorsal fins. 

 Here may be found numerous instances of shorter spines and rays from two- 

 thirds to one-fourth the normal length. But among so many sjiecimens it is en- 

 tirely probable that these spines and rays would be found in every possible stage 

 of growth. But this is not the case. The spines and rays, although sometimes 

 only one-fourth the full length, are always strong and suggest aborted rather than 

 immature structures. Besides, if this were the case, we would expect to find the 

 tendency toward a lower number of spines, and rays very decided in the first 

 year specimens. While this condition is true in the dorsal and anal rays, it is; 

 decidedly not true in the dorsal spines, where the characters in the first years are 

 almost identical with those of the third year. 



Natural Selection. — The principle of natural selection, the infiuence of 

 which upon this species I hoped in the onset of this work to find, can not be 

 applied in explanation of the difference in the number of scales and dorsal spines 

 without serious objections. If natural selection were the determining factor in 

 producing these differences, we should expect all the variations graduated with 

 the age. We would expect to have a narrower range of variation as the specimens 



