of the FUhcry Board for Scotland. 



207 



quently there is no reason why, if we form two or three races, we may not, 

 from these data, form a dozen. If we separate, however, the summer from 

 the winter fish, and examine them in this light, we do in some respects find 

 reason to think such a racial distinction may be a fact. Where the difference 

 between the summer and winter fish is not at all or only slightly marked, 

 as in the case of the head, we find, as of course might be expected, that 

 this occurs in similar combination with the other characters, both in the 

 winter and summer fish. But where we find a distinction between the 

 summer and winter fish in respect to any two characters, it will generally 

 be found that the proportional form which the combination takes among 

 the winter is held also among the summer, although of course not neces- 

 sarily in the same actual condition — a ground, I think, rather for doubting 

 their racial distinction than otherwise. The examination then of a large 

 number of the different combinations in which two characters are found on 

 the individual fish, leads me to the conclusion that no racial distinction is 

 shown by these alone, i.e., when the season in which they were caught is 

 ignored, for there is a gradation towards each extreme of the variations 

 which connects all the fish together. This is more marked the greater we 

 make the number of divisions indicating the variations. When we come 

 to a combination of more than two characters, the third character is found 

 so scattered throughout the combinations of the others that no value can 

 be placed on it. Taking all the fish, both summer and winter, and tabu- 

 lating them according to the combination of the dorsal and anal fin positions 

 as in Table XXXI., we see that from those with the most anterior dorsal 

 combined with the most anterior anal position, there is a gradation towards 

 those on which the median position is represented, and from them to those 

 fish with both fins in their most backward position. This, while it may be 

 due to variation not altogether irregular but altering according to the 

 growth of the fish, may also indicate a racial distinction, the line of de- 

 marcation between which it would be difficult to fix, on account of the 

 comparative regularity of the variation, and of the great predominance of 

 the median position. We apply then another combination, say the position 

 of the pectoral fin. If there is racial distinction we may expect to find 

 certain conditions of the pectoral fin combining pretty regularly with 

 certain definite combinations of the dorsal and anal, while the other condi- 

 tions of the pectoral would be found present with separate and distinct 

 grounds of dorsal and anal variation. At least, we should expect to find 

 most, if not all those fish which possessed the pectoral fin of one extreme of 

 variation, distinctly separated in the position of their dorsal and anal fins 

 from those which possessed the pectoral fin in its opposite extreme position. 

 If not, we can only conclude that the pectoral fin varies indiscriminately 

 and irrespective of the condition of the dorsal and anal, and is, therefore, 

 not of racial value. Such, in the herrings I have examined, is actually 

 found to be the case. Dividing the ground of variation of the position of 

 the pectoral fin into four classes (E have done it with eight classes, which 

 gives similar but even clearer evidence), it is found that those fish in the 

 first division, i.e., with the fin in its most anterior position on the body, 

 present all forms of combination of the position of dorsal and anal fins, 

 from those which have the dorsal and anal fins in their most extreme 

 anterior position to those which have them in the most posterior. The 

 very same remark applies to those in the fourth division with the pectoral 

 fin in its most posterior position. This is the more remarkable since very 

 few of the fish, comparatively, have the fin in these extreme forms. A 

 modified example of this follows. For instance if we divide the amount 

 of variation of the dorsal, anal, and pectoral fin positions into two moieties 



