1897.] On the Characters of Macropetalichthys. 497 
continuity backward, gradually tapering as far as the hinder 
margin of the dorsal plate. That this is a coherent element 
by itself is proved by its occurrence in the detached condi- 
tion; of several examples that have been met with, one is 
represented in Pl. XII, fig. 2. To distinguish it from other 
structures it may be called the dorso-central. On either side of 
the dorso-central is an apron-like expansion, which covers the 
remainder of the dorsal plate as far as its lateral margin, and 
incidentally conceals the posterior cranial angles. It is united 
with the dorso-central by sutures, and also with the plates in 
front of it in like manner. It, too, occurs in the detached 
condition, as is shown in fig. 3 (DL), and is worthy of being dis- 
tinguished by a separate name. We propose to call it the 
dorso-lateral. 
From the foregoing, it will be seen that portions of the 
cranium and also the dorsal element, to be regarded as a unit 
by itself, are covered by superficial scutes which are incapable 
of correlation with derm plates in other Arthrodires. Possibly 
they were originally coextensive with the underlying elements, 
which are now greatly reduced in thickness; afterwards they 
may have coalesced with one another and acquired a simpler, 
although still symmetrical arrangement. This condition is 
unparalleled among fishes, but a resemblance to it may be 
found in the corneous layer covering the carapace of tur- 
tles. The fact, however, that some of the tuberculated scutes 
no longer correspond with the subjacent elements, arouses a 
suspicion that the remaining scutes may present more or less 
modified conditions. It is to be noted that their arrangement 
differs from the pattern of most Arthrodires; but inasmuch as 
a separate system of plates has not been observed except in 
the posterior region, we may presume that in advance of this 
area each derm plate was covered by its own tuberculated 
coating. 
The problem of determining homologies among the cranial 
plates is by no means an easy one. It is certain that the plate 
termed for convenience the “central” corresponds to much 
more than the like-named element in other Coccosteids; in it 
are included at least the median occipital, the parietals and 
