THE VARIATION, RACES AND MIGRATIONS OF THE MACKEREL. 257 



of vertebrse, and equally invariable. This is obviously not the case. 

 But it is important to notice that the average, or rather modal, nuniber 

 of bars is exactly the same as the number of vertebrae. The modal 

 number of bars between shoulder girdle and last dorsal fin let is shown 

 in my tables to be 27. The total number of vertebras in the mackerel 

 is 31. But the vertebral column extends a little in front and a little 

 behind the limits mentioned, and by dissection I have found that this 

 excess coincides with the extent of the two anterior and the two 

 posterior vertebrae. Consequently the number of vertebne in the 

 region corresponding to the bars enumerated is four less than the 

 total number, viz., 27, which is also the modal, or most frequent 

 number of bars. This correspondence furnishes a conclusive con- 

 firmation of the general accuracy of my data in regard to the present 

 character, as well as of my statement that the bars are formed in 

 relation to the myotomes. The variation in the number of bars is 

 due to the fact that the pigment is not deposited in a regular manner 

 along the free surface of the myotomes, except in very rare cases. In 

 most fishes the pigment streak can be seen to correspond with a 

 particular myotome for part of its extent, and then to become 

 broken or discontinuous, either remaining as an isolated fragment, or 

 more frequently effecting an abrupt anastomosis with the pigment 

 of a neighbouring myotome. These two facts, the discontinuous 

 deposition of pigment along the myotomes, and the irregular 

 anastomosis of the pigment streaks of neighbouring myotomes, account 

 not only for the variable shape of the transverse bars, but also for 

 their variable numbers. The variation in the number of the bars, 

 in spite of the fundamental relation between the bars and the 

 myotomes, does not involve any variation in the number of the 

 myotomes or vertebne. But, just as the number of myotomes is 

 fixed for each fish, so the number of pigment-bars which are formed 

 in connection with them is not subject to alteration after the period 

 of formation. This period falls between the metamorphosis of the 

 larva and the acquisition of a length of some 4 or 5 inches.* 



To remove all doubt concerning this important point, I have calcu- 

 lated the mean number of bars in all the very small fish taken at 

 Plymouth during the autumn of 1897. The total number of fish 

 was 67, ranging in length from 5^ to 10 inches. The mean number 



* It is curious to notice that the transverse bars of the adult mackerel occupy a position 

 in relation to the myotomes which is quite diderent from that occupied by the vertical 

 lines of chromatophores in young mackerel from 14 to 18 mm. in leugth, according 

 to Holt's description and figures (this Jourual, V., 1898, p. 110, figures 3 and 4). In 

 preserved specimens at this early stage Holt states that, ' ' on the sides of the trunk, the 

 chromatophores are set more thickly at the lines of division of the myomeres than 

 elsewhere," i.e., along, instead of between, the septa which separate the myotomes. 



