148 Dr. R. H. Traquair—Devonian Fishes from Canada. 
The anterior median ventral is not here exhibited, but as it occur 
detached in another specimen I have in the figure inserted it in its: 
place. ‘The median ventral is however well shown in position, and 
detached examples of it occur showing both outer and inner aspects. 
Both of these median plates closely resemble the corresponding 
plates in Coccosteus. 
So does the anterior ventro-lateral (a. v. 1.) which is quite cor- 
rectly identified by Whiteaves (op. cit. pl. ix. fig. 3). Now attached 
to the antero-external margin of this plate is seen the spine-like 
body recently described by Mr. Smith Woodward. This spine 
occurs not only in the specimen on which the above restored sketch 
is founded, but in the collection now before me is found both abso- 
lutely detached, and also in articulation with otherwise isolated 
examples of the anterior ventro-lateral plate. 
Philyctenaspis Acadica, Whiteaves ; restoration of ventral carapace. a@.v./. anterior 
ventro-lateral plates; p.v./. posterior ventro-lateral, not defined in the outer 
margin in the fossil; @.m.v. anterior median ventral; m.v. median ventral ; 
i... interlateral; c. cornu or lateral spine. 
Mr. Woodward mentions these spines as “‘ each apposed to a lateral 
plate of the trunk :” it is now shown to which lateral plate, as well as 
to which part of the dermal armour of Coccosteus, they correspond. 
It will be seen that the spine (c) touches with its proximal 
extremity a narrow portion of bone (7. J.) which is articulated with 
the anterior margin of the anterior ventro-lateral plate (a. v.1.). On 
turning to my previous restoration of the ventral cuirass of Coccosteus 
decipiens, Ag.' it will at once be seen that this portion of bone cor- 
responds exactly to the anterior-inferior part of the interlateral plate, 
(7. l.) which occupies the same position with regard to the anterior 
ventro-lateral, and meets its fellow of the opposite side in front of 
the anterior median ventral (a. m. v.) while the spine in like manner 
corresponds to the short process which passes backwards separating 
the anterior ventro-lateral below from the antero-lateral above. 
This process, short in C. decipiens, is more largely developed in C. 
minor, H. Miller, in which species I have stated that it has a “‘ very 
spine-like appearance,” though owing to the proportionally greater 
_1 Ann, and Mag. Nat. Hist (6) vol. vy. 1890, pl. x. fig. 3. 
