472 ALLELES. {Vou.H: 
eye, it somewhat more than half encircles the orbit, extending 
to a point above and behind it, and a little in front of the blind 
upper end of the spiracular canal. The canal here turns sharply 
backward, and, as the squamosal or temporal portion, is continued 
backward and upward above the pre-opercular fold, to a point 
between this fold and the upper end of the opercular opening, 
where it gives off the supratemporal cross-commissure. The 
post-temporal portion lies behind this commissure, continuing 
at first in the line of the squamosal canal upward and backward, 
above the upper end of the opercular opening, and then down- 
ward and backward under the upper and posterior margin of 
the operculum, to the hind edge of the supraclavicula, where 
it joins the anterior end of the lateral canal of the body. 
2. Supra-orbital Canal.— The supra-orbital canal begins a 
little median to and behind the nasal tube. It runs at first 
toward the median line, and then almost directly backward 
above the eye, ending near the hind margin of the frontal, 
and sending its posterior branches into the anterior part of 
the parietal and squamosal. Behind the eye it is deflected 
somewhat laterally, and anastomoses with the infra-orbital canal 
at the bend in that line where the suborbital portion joins 
the squamosal. The arrangement of the canals in the adult 
at this point is such that the supra-orbital has always been 
considered the direct continuation forward of the squamosal 
or temporal portion of the infra-orbital, the two together being 
described as the main lateral canal of the head, and the suborbi- 
tal as one of its branches. The development of the canals 
in Amia shows conclusively that this interpretation is wrong, 
for the supra-orbital develops independently, its innervation is 
different, and it only acquires its connection with the main 
infra-orbital as an anastomosis after both canals have been fully 
inclosed. 
That this independence of the supra-orbital canal is not 
peculiar to Amia is shown by the arrangement in the Chara- 
cinidz as given by Sagemehl (No. 14, p. 36). In these fishes 
the supra-orbital has, in the adult, the condition found in the 
larva of Amia; that is, it is separate and distinct from the main 
infra-orbital, broken off from it, as Sagemehl says, by the in- 
trusion of the anterior end of the dilator operculi muscle, 
which has in this fish an unusual insertion on the upper sur- 
