Nos. IAND2.] ANATOMY OF SCOMBER SCOMBER. 175 



arising from the lateral aspect of the canal. It ran at first for- 

 ward and slightly laterally, almost parallel to the main canal, then 

 turned sharply laterally, and, branching dichotomously, ended in 

 two pores which lay directly between, or one of them even slightly 

 in front of, the exposed portions of the eyeballs. 



The sixth system lies on the mesial side of the main canal and 

 was always found as a group of two, three or four tubes, which 

 arose separately and independently from the main canal, just at 

 the bend in that canal and close to the point of origin of the trunk 

 of system 5. Where there were four tubes the anterior tube of 

 the group ran mesially and forward, the second tube mesially, the 

 other two mesially and backward. Each tube ends on the outer 

 surface of the head in a single pore, the four pores, in the speci- 

 men used for illustration, forming a nearly straight line, lying 

 close to the middle line of the head, and close to the corresponding 

 line of the opposite side. There is thus, in Scomber, an approach 

 to a supraorbital, or frontal commissure. 



The trunk of the next system, system 7, formed, in most of the 

 specimens examined, a direct continuation of the main canal. It 

 had every appearance of being the terminal section of that canal, 

 and, in all the earlier dissections, was considered as such. The 

 arrangement found in the corresponding part of the lateral system 

 of Anna strongly suggesting that the main canal would not nat- 

 urally so end, repeated dissections were made in order to establish 

 what I considered as the true relations. These were finally found, 

 as shown in the several figures, and were found repeatedly after- 

 wards. A somewhat different arrangement, shown in Fig. la, 

 has however since been found on both sides of the head of three 

 large specimens, by ]Mr. Nomura. Whether this represents a 

 difference due to age, or to some difference in species, or simply 

 a variation in the formation of the canals, I am unable to judge. 

 In the arrangement considered as the normal and proper one, the 

 penultimate system of the supraorbital canal anastomoses, ex- 

 actly as in Auria, with that system of the main infraorbital canal 

 that is formed at the postorbital bend in the canal, and a double 

 system arises. This system is, accordingly, in Scomber, system 

 II infraorbital — 7 supraorbital, as already stated in describing 

 the main infraorbital canal. In the subdivision of this double sys- 



