EASTMAN: THE DINICHTHYIDS. 39 



The first of these (Plate 3, Fig. 4) is easily recognizable as the left antero- 

 dorso-lateral of a new species of Dinichthys, and is chiefly remarkable for its 

 finely tuberculated style of ornament. This plate is nearly twice the size of the 

 corresponding element described by Newberry as D. tuberculatus, its articulating 

 condyle is differently situated, and the tuberculation is entirely dissimilar. 

 Of D. tubsrculatus, Newberry ^ speaks as follows : " The tuberculation of the 

 surface is relatively coarse, and the tubercles vary much in size and are irregu- 

 larly scattered. Most of them seem to be hemispherical and plain, but others 

 are more or less pitted, and a few are stellate." In the present species the 

 tubercles are small and closely crowded, and are distinctly stellate at their bases. ^ 

 It is somewhat surprising that there should be so few American species which 

 present the characteristic surface ornamentation of the Coccosteidce ; the infer- 

 ence is that the tuberculated are more primitive than non-tuberculated forms. 



A longitudinal fracture traverses the plate to the left of the sensory canal. 

 It is interesting in that it displays very clearly the course of the vascular 

 (Haversian) canals, which run essentially parallel with the surface of the plate. 

 The canals are also well shown where the articulating condyle has broken off ; 

 and from their direction it would appear that the plate had grown by incre- 

 ments to the visceral surface only. 



The second specimen in this collection that deserves notice is evidently the 

 impression of one of the ventral plates, probably the left antero-ventro-lateral, 

 the substance of the bone itself being entirely worn away. The surface orna- 

 ment cannot be discovered from this specimen, but several fragments associated 

 with it exhibit the same tuberculation as occurs on the autero-dorso-lateral just 

 described. The only reason for disassociating the two specimens specifically is 

 that they represent individuals of somewhat different size ; but the dispropor- 

 tion does not appear of itself sufficient ground for separation. The supposed 

 antero-ventro-lateral measures 23 cm. in length by 11 cm. in wadth at about the 

 middle of the plate. How much of the anterior portion is wanting cannot be 

 accurately determined. Another large specimen from the same locality is 

 to be seen on exhibition in the United States National Museum, at Washing- 

 ton, D. C, bearing the catalogue number 14,821. 



Fragments of various size, and indistinguishable from this species so far as 

 one may judge from the ornamentation, have been collected by the writer in 

 the State Quarry fish-bed, near North Lilierty, Iowa.3 Other remains have 

 been found in the Cedar Valley Limestone of the same State by Professor 

 Samuel Calvin. One of the largest of these, which belongs to the State Uni- 



1 Newberry, J. S., Palaeozoic Fishes of North America (Monogr. U S. Geol. Surv., 

 Vol. XVI. p. 99), 1889. 



2 The artist has represented these somewhat diagrammatically in Figure 4, with 

 the result of imparting a rougher aspect to the plate than is natural, although it is 

 plain that the original has suffered somewhat from abrasion. 



* See notes "On the Occurrence of Fossil Fishes in the Devonian of Iowa," 

 appended to Report on the Geology of Johnson County (pp. 108-116), by Samuel 

 Calvin, State Geologist. 1897. 



