THE VAEIATIOX, RACES AND MIGRATIONS OF THE MACKEREL. 263 



The figures in the table completely disprove the accuracy of 

 Donovan's statement, and show, once for all, that the shape of the 

 bars is entirely independent of the sex of the fish. 



The percentages of the total numbers of males and females examined 

 are 43 ^ and 57 $ ; and for the first two types of bar this proportion 

 is almost exactly reproduced, being 45:55 for the geniculate bars, 

 and 46:54 for the wavy bars. So far from the straight bars being 

 distinctive of male fish, there is a slight preponderance of females 

 showing this character, the proportion being 36 (? to 64 $ . For a total 

 of 72 fish, however, this proportion is sutticiently close to the total 

 proportion of males to females to show the entire absence of sexual 

 peculiarities in the matter. 



VIII. Dorso-Lat£iial Intermediate Spots. 



The general character of these spots has already been given in the 

 description of the characters investigated (p. 241). They are, strictly 

 speaking, discontinuous portions of the pigment streaks of the 

 myotonies, but they assume so definite a shape, and possess such 

 clearly-defined relations to the transverse bars, that it is possible to 

 discuss the variations in their frequency independently of the bars of 

 which, theoretically, they form a part. 



No fragments of the bars are here considered as " intermediate 

 spots " unless they possess a sharply-defined round or elliptical shape, 

 and unless they are situated between two neighbouring transverse bars, 

 or are entirely or partly enclosed in a ring-like modification of two 

 neighbouring bars. 



It sometimes happens that the extremities of the bars are separated 

 off as small " end-pieces," which may even assume a rounded form. 

 Such end-pieces are undoubtedly connecting links, from a theoretical 

 point of view, between perfectly linear and continuous bars and the 

 pigment-spots which I distinguish as " intermediate " ; but if these 

 end-pieces plainly continue the lines of the bars, and do not occupy 

 an isolated intermediate position between two adjacent bars, they have 

 been excluded from consideration. 



The importance which I attach to the extreme type of " intermediate 

 spots," as defined, rests on the following grounds. The shape and 

 course of the bars is fundamentally determined, as already shown, by 

 the shape and course of the myotomes or muscle-segments. Bars, 

 therefore, which follow the course of the myotomes may be regarded as 

 primitive in character. In such cases, since the surface of every 

 myotome is occupied by a pigment-streak, it is clear that spots, having 

 an intermediate position between two adjacent bars, can have no 



