Dy. EL Traquair—Homosteus and Ooccosteus compared. 3 
fig. 1), he will see that this specimen at least, far from having “no 
reference to Pterichthys,” is the median occipital plate of a very 
closely allied form indeed. Without injustice to the memory of Sir 
Philip Egerton, to whom paleichthyological science is indebted for 
so much valuable work, it is clear that he had not sufficiently gone 
into this question at least, as the fact above noted was dwelt upon 
by Pander himself (op. cit. p. 16). 
On the Continent, however, the name Homosteus, Asm., has been 
adopted for Hugh Miller’s fish, and reasons have also been found 
for maintaining Pterichthys and Asterolepis as distinct genera, in a 
supposed difference in the articulation of the arms. In the previous 
volume of this Magazin! J have shown that this supposed diagnostic 
mark is untenable, at the same time that I have sought to establish 
another on the mode of articulation of the anterior median dorsal 
plate. 
But Agassiz’s work, in which he classified ‘“ Asterolepis”” among 
his “ Coelacanthi,” a group generally equal to the Cycliferous Crosso- 
pterygii of more recent times, was the means of drawing Hugh 
Miller into mistakes of much greater importance than mere nomen- 
clature. Accordingly, as Pander pointed out, Miller attributed to 
_his Asterolepis “the teeth of Dendrodus and the scales of Glypto- 
lepis,” and made a very formidable creature out of it, ten to thirteen 
feet in length; indeed, referring one of the large Russian plates 
(Heterosteus, Asmuss) to the same genus, he calculated a length of 
eighteen to twenty-three feet for the entire fish. And his non- 
recognition of the true affinities of the creature led also to other 
mistakes in the identification of bones, to which allusion will be 
made in due course. 
By Asmuss Homosteus and Heterosteus were placed in a family by 
themselves, Chelonichthyda, in a somewhat heterogeneous group of 
“‘ Ganoidea loricata,” the other families herein included being Spatu- 
larida, Acipenserida, Coccosteida, Pterichthyda, and Cephalaspida. 
As regards Homosteus, though, as Pander remarks, it is wonderful 
how, without knowledge of Hugh Miller’s drawings or description, 
he was able to fit together the isolated plates at his disposal, yet, 
unacquainted with the orbits, he supposed the cuirass to belong ex- 
clusively to the body, and also entirely reversed its position on the 
animal. 
Pander, however, classified Momosteus in McCoy’s group of 
Placodermata and rightly gave it a place immediately after Coc- 
costeus, interpreting as its median dorsal plate the one considered by 
Hugh Miller to be a ‘“‘hyoid,” and supposed by him to occupy a 
place between the rami of the jaws. This supposed hyoid plate 
was known to Hugh Miller only in an isolated form, but it fell to 
the lot of the late Mr. John Miller, of Thurso, to record a specimen 
in which it occurred in its natural position on the back behind the 
head, and so with absolute certainly to confirm Pander’s view of the 
case. Mr. Jobn Miller’s collection having some years ago passed 
1 Notes on the Nomenclature of the Fishes of the Old Red Sandstone, Grou. 
Mae. Dee. III. Vol. V. 1888, p. 508. 
