EASTMAN: THE DINICHTHYIDS. 81 
of parts as known in Dinichthys does not hold true by any means for 
other genera (Trachosteus, Mylostoma, ete.) belonging to the same fam- 
ily; and this fact admonishes us not to press hypothetical correlations 
too far, even within specific limits. Caution is enjoined in this particu- 
lar case by yet another consideration. From the same locality and 
formation, Mr. Mixer has obtained a pair of mandibles associated with 
fragmentary Dinichthyid plates. The condition of these remains docs 
not warrant a precise specific determination, but their affinities are 
probably with D. minor. The length of each ramus is about 17 cm., 
and the maximum height 5 cm. Either, therefore, these remains and 
the plastron represent together but a single species (D. % minor), or 
we have evidence of two medium-sized species (D.? minor and D.? new- 
berryi) in the Portage Shale. 
Under these circumstances it is apparent that a positive identification 
of the species is impossible. For the sake of convenience, we might 
follow Dr. Clarke’s example, and refer all the detached plates occurring 
in the Genesee Shales to D. newberryi, and all those from Portage Shale 
to D. minor. But there is no reason for supposing that each of these 
horizons contains but a solitary species; the indications point rather to 
the presence of more than one species in both horizons. And there is 
no reason why the doctrine of correlation of parts should not be applied 
to all the species of Dinichthys until experience has shown it to be in- 
valid for some of them. Provisionally, therefore, we are in favor of refer- 
ring the Portage plastron to the species with which it most closely agrees 
in measurement and geological horizon, that is to say, with D. newberryi. 
On the other hand, the Portage mandibles that have just been men- 
tioned, and the detached dorso-median plates from the Genesee, may be 
referred provisionally to D. minor. 
Comparative measurements of certain derm-plates for several species 
of Dinichthys are exhibited by the table on the following page. 
Besides the plastron just described, there are several other interesting 
structures preserved on the same slab. In advance of the plastron are 
a number of badly weathered fragments, which evidently represent the 
dorsal plates of the body. The forward portions of both antero-ventro- 
laterals are covered over, and their proper boundaries obscured, by some 
of these fragments; but none of them are identifiable with certainty 
unless it be the antero-lateral tip of the dorso-median, which rests upon 
the angle of the right antero-ventro-lateral. This concealment of the 
underlying plates along their margins is unfortunate, since the restored 
anterior boundary has not such a clear basis of fact as one could wish 
