The Structure of Dinichthys 



135 



Museum in 1930 found, among compara' 

 tively good material of Coccosteus decipy 

 ens Ag., a head (No. E1257) which shows 

 a doubtful element at the extreme angle 

 of the head (Text'figure 19). In all proba' 

 bility it is PM. However, it is not very 

 surprising that PM has never been found 

 and described in Coccosteus. It is very 

 small, only a few millimetres long, and 

 difficult to see. 



Even in Dinichthys, where PM is 5 

 to 6 cm. long, it has never been observed 

 befoie. What is especially remarkable is 

 that this plate does not occur infrequently. 



Text'figure 18. 



The left hind corner of the head of Dinichthys inter- 



medius Nwb., showing the relative positions of PM, 



M and £0. 



In the collections of the American Museum 

 alone, it is to be found in three head roofs, in addition there are 5 or 6 isolated plates 

 in perfect preservation. As a curiosity, it might be mentioned that one of these plates 

 is labelled in Newberry's hand, ''New genus and sp." 



Only in Branson's reconstruction of Dinichthys terreUi Nwb. (1908.1, .2, 1911) do 

 we find some traces of this plate. As has been mentioned, in this reconstruction (Text' 

 figure 6) there is shown a new sensory canal, which goes nearly across to the outward 

 branch of the M sensory canal (Text'figure 12 V). It is enough to compare Branson's 

 figure with mine to be sure that Branson was mistaken and has taken a suture between 

 M and PM for a sensory canal (Text'figures 9 and 12). 



Central. — This plate (Text'figures 12, 13, 14 and 15 C) is the largest one of the 

 whole roof, and is placed in the central part of the head. Along the median line it meets 

 its corresponding plate from the other side. On the inside, C is naturally divided into 

 two parts: one, the hind, thick; the other, the fore, thin. The posterior part is nothing 

 other than the "connecting bridge" mentioned before. The thin part forms the bottom 

 of the posterior part of the neurocranial impression. On this thinner part, the ossifica' 

 tion rays are especially clearly developed. They radiate from the ossification center 

 placed nearly in the middle of the connecting bridge (CR). The limit between the thin 

 radiate and the thick smooth part of C is very sharp. This circumstance explains why 

 Newberry (1889), Claypole (1892.2), and Eastman (1897-2, .3) have described C as two 

 separate plates: "parietal" and "frontal." 



On the outside of C we find some sensory canals. Their courses and lengths vary 

 strongly, not only in different species, of Dinichthys but also in the different individuals 

 of the same species. Most usually there are two sensory canals which cross the upper 

 limit of C — one from PrO (Text'figure 12 I), the other from PtO (Text'figure 12 II). 

 They run downward side by side and nearly fuse in the center of the plate (ossification 

 center). In some individuals they nearly meet, in others one of these canals is longer and 

 continues farther down. This may in some species cross the hinder Umit of C. and con' 



