The Structure of Dinichthys 



141 



away from it. It curves forward and continues onto the handle (Text-figures 22 and 23 Rj) 

 where it runs nearly to the point of the tongue-shaped part. In the beginning this ridge 

 is low but rises quickly and reaches its greatest height about the curving point (Text- 

 figures 22 and 23 y). It keeps this height to the end. As seen from Text-figures 22 and 23, 

 it runs in a half-circle around the orbital impression and forms its lower wall. 



Another ridge begins at about the middle of the hind margin (Text-figures 22 and 23 

 R.j). Quickly rising, it runs forward to the curving point of the first ridge (Text-figures 

 22 and 23 y) where it makes a sharp curve, 

 and forms a high crest (Text-figures 22 

 and 23 Pr). From this point the tongue- 

 shaped part of the handle begins. As we 

 know, the tongue-shaped part is placed 

 behind the outside of SO and thus its in- 

 ner surface is nearly on the high plane of 

 the ridges. After a sharp curve, this 

 ridge runs along the middle of the tongue- 

 shaped part, but now as quite a low edge 

 (Text-figures 22 and 23 R4). 



Between the orbital wall and this 

 edge runs a groove (Text-figure 22 gr). It 

 is deeper and narrower at its beginning, 

 but becomes flatter and broader forward. 

 A very clear, deep impression (Text-fig- 

 ures 22 and 23 ip), located between the curving point of both ridges, connects it with 

 the posterior part of SO. As we see, the structure of SO is very complicated and difficult 

 to describe. It can be better understood by studying Text-figures 21, 22 and 23. We 

 may point out furthermore that the outline of SO is rather variable in different species 

 . of Dmichthys. Thus in Dmichthys terrelli Nwb. its blade is relatively long and narrow, 

 and the handle shorter; in Dinichthys interynedius Nwb., on the contrary, the blade is 

 broad and relatively short, but the handle is strong and massive (Text-figure 24). 



Post-Nasal. — This plate, which is represented only by very imperfect fragments in 

 the collection in the American Museum, has never before been described or figured in 

 detail, but its presence in other Arthrodira is well known. Traquair (1890.4) for the first 

 time figured it in Coccosteus and called it pre-maxillary (he called the SO the maxillary). 

 He correctly marked the nostril openings as an impression on this plate. Woodward 

 (1892.1, .2) described the same plate in Phlyctaenaspis acadica Wh., but did not find any 

 clear trace of the nostrils. 



Stensio in 1925 based his reconstruction of Phlyctaenaspis on the same specimen as 

 Woodward. He called the pre-nasal ''plate L" and let ''V together with PtO bound the 

 orbits on the upper side. In this manner PrO was surrounded by C, PtO, P7v{, P and R, 

 and, contrary to what is found in all other Arthrodira, was excluded from bounding the 



Text-figure 24. 



The sub-orbital plates of (A) Dinichthys terrelli Nwb. 

 and (B) of Dinichthys intermedms Nwb. drawn to show 

 the relative siz;es of the fore and hind part of the plates. 



