Structure and Classification of the Asterolepidse. 489 



Bodi/ -carapace. — This, as already shown by Sir Philip 

 Egerton and Hugh Miller, is nearly quite flat below, high 

 and vaulted above, the sides rising at right angles to the 

 base ; as the lormer author says, " the contour must have had 

 considerable resemblance to a high-backed tortoise with the 

 carapace cuhninating near the anterior margin," the trans- 

 verse section being " not unlike the outline ot a stirrup-iron." 

 It is composed altogether of thirteen plates, being two more 

 than the number given by Hugh Miller, but agreeing in this 

 respect with Asterolejns as described by Pander. 



The general I'orra of these plates is already so well known 

 from the descriptions of Hugh Miller and Egerton that I 

 need here only allude to certain matters of detail which 

 require correction, some of them, however, being of consider- 

 able importance. In PI. XVII. tigs. 1, 2, and 3, 1 have 

 represented the outlines of the body- plates as seen from the 

 back, belly, and side respectively, the thick black lines repre- 

 senting overlapping edges, as seen on the external surface, 

 the thnr ones those which are overlapped, and which conse- 

 quently are concealed externally when tiie plates are in situ. 



The tirst point of importance is the presence of two small 

 narrow plates {s. I. tig. 2), each of which occupies a space cut 

 out from the inner half of the anterior margin of the anterior 

 ventro-lateral (a. v. J.) and is in contact mesially with its 

 fellow of the opposite side. This is Pander's semilunar 

 in Asterulejjts (7, pi. vi. tig. 1), and though not men- 

 tioned in the descriptions of Hugh Miller and Sir Pliilip 

 Egerton, the space which it occupied is distinctly seen in one 

 of Sir Piiilip's tigures (8, p. '605, tig. 2) . 



jSext, as to tlie anterior ventro-lateral plate itself and the 

 mode ot articulation of the arms. Notwithstanding the con- 

 trary opinion of Hugh Aliller and M'Coy, Sir Philip Egerton 

 strongly maintained that the arms were articulated to separate 

 "thoracic" plates, marked off by a distinct line of suture 

 from the anterior ventro-lateral ; and so contident was he in 

 this opinion that he went so far as to say that he was " at a 

 loss to conceive how Pioiessor Pander can have been led to 

 assign the attachment ot the arms to the ventro-lateral plates 

 as shown in the magnitied tigure on tab. vi. of his magniti- 

 cent work on the Devonian tishes, althougli in the preceding 

 piate these organs are correctly drawn as appended to the 

 ihoKicic plate" [\), p. 105). JSIow in this matter Pander-'s 

 accuracy cannot be impugned as far as Asteroiejjis is con- 

 cerned, tor the Kussian plates of this genus were found iso- 

 kiteU and uncompressed, and the place of articulation of the 

 arm can easily be veiitied on a specimen of the anterior 



