758 THE AMERICAN NATURALIST. (VOL. XXXII. 
and central elements are not nearly so heavy as in other species 
of Dinichthys, and the formation of the double socket in the 
middle of the nuchal line is simpler. This peculiar structure 
is supposed by Claypole to mark “the place of insertion of 
some powerful muscle or ligament that connected the head 
with the rest of. the body.” It is well shown in a number of 
Newberry’s figures of Dinichthys and Titanichthys.1 The 
bone is extremely dense in this region, and the thickness of the 
cranium is greater than in any other place. Hence, fragments 
that have been rolled about or subjected to weathering often 
become reduced so as to leave nothing but this portion of the 
occiput. 
Dinichthys terrelli Newberry. 
This species is numerically the most abundant of American 
Dinichthyids, and the largest collection of its remains is pre- 
served in Columbia University.? The writer, having but one 
head at his disposal in the Agassiz Museum, has not essayed 
to figure the cranial osteology, but we may say it does not differ 
materially from that of D. intermedius. The Cambridge speci- 
men shows no evidence of a division of the central into two 
parts, but the solidifying ridges on the under surface are de- 
tached from the cranial bones for a slight distance anteriorly, 
thus producing the semblance of separate plates. 
1 Loc; cit. (1889), Pl. I, Fig. 2; Pl. IV, Fig. 2; PL VIII, Fig. 4; Pl. LII, 
ig. I. 
2 Since this article was written the writer has enjoyed the privilege, thanks to 
the courtesy of his friend Dr. Bashford Dean, of looking over the greater part 
of Professor petir by s collection, which has recently been stored in cases in 
Schemerhorn Hall. specimens could be found to prove the existence of 
“parietal ” and “ wate ” elements, and the conclusion is that they do not occur. 
A large example of the antero-dorso-lateral with entire margins (embe bedded in 
shale) proves that this plate extended underneath the dorso-median and clavicular 
elements for a distance hitherto e, the covered area being even greater 
than the exposed. Newberry’s figures of the clavicular in this species are seen to 
be based upon an imperfect specimen, the superior margin of which has been arti- 
ficially restored, and is to a certain extent misleading. On the other hand, the 
collection contains some unusually perfect examples of this plate, which certainly 
PAAA to be figured. There is also abundant evidence to show that the normal 
condition of the ventro-median plates in D. ¢erre//i was one of simple overlap, 
but, in the adult stage, fusion of the two elements may progressively set in 
