52 
lateral edge of the parietal becomes “so weit verkürzt” that practically 
no roof whatever is left to the hole. The same conditions are said 
to be found in many of the Physostomi, and in nearly all of the 
Acanthopterygii and Anacanthini. In Sclerognathus the hole is said 
to be reduced to “eine tiefe, von medial und hinten von scharfen 
Rändern begrenzte Grube”. Sclerognathus thus seems to present ex- 
actly the conditions found in Scomber, and to which I have already 
referred in an earlier work (No. 4, p. 92). I there said: “This seems 
to indicate that Amia and Scomber represent separate lines of descent 
from some fish in which the trunk muscles had not as yet invaded 
the temporal part of the skull to the extent they have in these two 
fishes.” A similar conclusion is perhaps indicated in SAGEMEHL’s 
statement (No. 29, p. 551): “Selbstverständlich sind die bei diesen 
Formen (Cyprinidae) und bei Sclerognathus so ähnlichen Bildungs- 
verhältnisse ganz unabhängig von einander entstanden zu denken.” 
That the condition found in Scomber is not due simply to the dis- 
appearance of the plate of bone that forms the dorsal, dermal part 
of the squamosal of the Characinidae and Cyprinidae, and the whole 
of the bone in Amia, seems shown conclusively by the fact that the 
hind end of the supraorbital canal lateral, which, in all these fishes, 
lies in the frontal bone, lies, in Scomber, internal to the trunk muscles 
that fill the temporal groove, while, in all the others, it lies morpho- 
logically external to those muscles. 
From the hind end of the squamosal, both in the Characinidae 
and Cyprinidae, a process, usually long, is said to project down- 
ward and backward and to serve for the attachment of the supra- 
clavicular. 
According to SCHMID-MONNARD (No. 30) the squamosal of Tele- 
osts is always formed of at least two different components, and there 
may be even three or four such components, all fused together to 
form the single bone of the adult fish. One of the two components 
that are said to be always present is partly of subperichondrial origin 
and partly of endochondrial origin. It is said to first appear as two 
thin layers of bone, called by ScHhmip-MonnArD the primary bony 
lamellae, both lying internal to the perichondrial membranes, one on 
the outside and the other on the inside of the cartilage that encloses 
the summit of the external semicircular canal. These two layers of bone 
are thus of exochondrial origin although subperichondrial in position, 
Internal to them the cartilage ossifies directly, true endochondrial bone 
thus arising, continuous with the two primary exochondrial lamellae 
The second component always found is, in young embryos, wholly 
