166 Bashford Dean Memorial Volume . 



distinct from the hinder thick part which is not overlapped. The upper margin (Text- 

 figure 49 MD) underlies the plate MD, the lower is under" ADL and PL (Text-figure 

 49 ADL, PL). The part overlapped by PL is sharply defined. It is more impressed 

 and shows deep grooves. The connection between PDL and PL becomes still more solid, 

 as the upper spine-like angle of PL sticks deeply into the margin of PDL (Text-figure 

 49 X, section a-h). 



On the part of PDL which is not overlapped runs a sensory canal (Text-figure 49 

 XII). It forms the immediate prolongation of the before described canal on ADL, and 

 runs slightly upward from the boundary line between these two plates to the ossifica- 

 tion center of PDL. When there are two or more parallel canals on ADL, we also find a 

 corresponding number of canals on PDL (Text-figure 55). On the inside PDL shows 

 nothing of special interest. Its surface is relatively smooth. Attention is merely called 

 to a groove-like impression, running from the middle of the hind margin obliquely down- 

 wards to the hind side margin of the plate (Text-figure 50 gr). 



Antero-Lateral. — This plate (Text-figures 51, 52, 53 and 55 AL) is the most 

 complicated one in the whole body carapace. It was well known by older investigators, 

 but its true outlines and divisions were misunderstood. 



Newberry in 1889 described the first complete AL and called it "clavicular."' This 

 plate was in reality not only AL, but a complex of three different plates — namely AL, 

 Sp and IL. In reconstruction they had been so strongly fastened together with plaster 

 that it was absolutely impossible to find the boundaries between these three components. 

 From Newberry's time on this "clavicular" was accepted as a classic and in later papers 

 and reconstructions it was considered a single plate. It is interesting to point out that as 

 early as 1893 Claypole (1893.1) had described and figured a clavicular plate from Dinich- 

 thys or from Titanichthys which shows very clearly a border dividing a short part of the 

 "clavicular" from the hind part. The front part is Sp; the hind, the real AL. But this 

 observation remained unnoticed. 



Later, various authors came to the opinion expressed by Eastman (1907) that 

 ". . . in Dinichthys the inter-lateral does not occur as a distinct plate, but may be repre- 

 sented by the fork-like prolongation of the so-called clavicular, the broad upper portion 

 of which corresponds of course to AL." However, Newberry's "clavicular" is not only 

 composed of three separate bones, but is also strongly deformed by pressure. Newberry 

 did not correct this mistake, but only filled up the clefts with plaster. As a result his. 

 "clavicular" becomes too flat and its inner wing runs nearly parallel with the outer. 

 This fault misled Dean in his reconstruction of Dinichthys (1909). Instead of combining 

 the inner wing of the clavicular with the ventral shield, he made it a supporter of the 

 infero-gnathal. We also find the same mistake made by Bryant (1918). 



In reality, the antero-lateral (Text-figures 51, 52 and 53) is a big plate with a large 

 upper part (A), a relatively small lower part (B), and an inner wing (C) which extends 

 obliquely inward. On the outside the plate is nearly plane and smooth. In its over- 

 lapping relations AL shows an exception to the usual rule. It is not overlapped by the 



