of the Fishery Board for Scotland. 



299 



dorsal, anal, and pectoral fins, the two former being in general more back- 

 ward, and the latter more anterior in the fish of summer than of winter, 

 however, strengthens the assumption. But the distinction is so small, and 

 so many fish are found in both seasons with these characters identical even 

 in some cases to the greatest extremes of variability, that it would not be 

 surprising to find further prolonged examination of the fish do away with 

 this apparent distinction as a constant seasonal occurrence. 



Apart from this question of race, however, the examination of these 

 herrings gives some interesting results in showing to what extent variations 

 in the characters are due to the size of the fish, — and the following tables 

 are prepared in order to show this. 



Of course these tables must be looked at in the way of ascertaining the 

 general tendency towards any particular condition of a character, and are 

 not to be expected to show an absolutely exact gradation in the particular 

 direction expected ; for as we have already seen there is an unequal amount 

 — in general not very great — of variation ia all the characters investigated, 

 which partially obscures the general rule applicable to each. This varia- 

 tion can only be expressed as individual variation, and may arise from 

 various circumstances, partly no doubt of heredity, as well as the environ- 

 ment especially during their development and early life, but may differ- 

 ently affect many fish developed and living under exactly the same circum- 

 stances, or even produced by the same parent. 



In the tables I have departed somewhat from the method of calculation 

 employed in my previous paper. In it the characters are generally com- 

 pared with the length of the fish, including the head length but excluding 

 the tail. In the following tables, however, for more correct comparison, I 

 have reduced all the measurements to a ratio of the length of the body 

 proper, excluding, for reasons previously stated, the head length as well as 

 that of the tail. The terminal point of the body is fixed as previously de- 

 scribed, and the length of head excluded is anterior to the posterior dorsal 

 ridge of the supra-occipital bone. I have already given a table indicating 

 the proportional variation of the head length to the total length, less the 

 caudal fin length, but no indication was given of the correctness of the 

 general statement there made, that the head did not increase in length at 

 an equal rate with increase in the body length ; a consequence of which is 

 that taking into account the mature (sexually) fish only, the shorter of 

 these are found as a rule to have the longer head. This is found to occur 

 in both the summer and winter herrings. 



If the figures showing this fact were artificially collected into one or two 

 sets of percentages, it might be supposed therefrom that possibly the dif- 

 ference between the head length in the larger and smaller fish indicated a 

 difference in race, but the greater the number of classes into which we 

 collect the fish for the purpose of comparison — i.e., the nearer we approach 

 to an individual examination of them — the more do we see how gradual is 

 the change in the ratio. Thus, if we put it down to racial causes, we must 

 suppose an absurdly large number of these. Moreover, an examination of 

 the circumstances under which the herrings exhibiting these varying char- 

 acters were procured, effectually disproves this, for in fish of all the varying 

 sizes procured at one and the same time and place, we find a corresponding 

 variation in the head length, and most of the other characters. It seems 

 quite clear, therefore — even without examination of the head of the imma- 

 ture fish, which shows a still less ratio to the body length than obtains 

 among the mature, although subject to the variation and overlapping found 

 to occur in nearly all the characters, — that this decrease in the proportional 

 length of the head to the length of body is a result of increase of size, and 



