OF THE AMEETCAN PTERASPIDIAN", PAL^ASPIS. 551 



existence of the genus Holasjpis, though it had been described in 

 1873 by Prof. Lankester, who kindly called my attention to it by 

 sending me a copy of his paper. Had I known this earlier I should 

 certainly have referred my species to his genus instead of coining a 

 new term. However, as it has since proved that Prof. Lankester's 

 name was pre-occupied, my own has precedence, and Holasjpis has 

 consequently been referred to Palceaspis by Mr. Smith Woodward in 

 his Catalogue. But Holaspis can only remain a congener of Palce- 

 aspis on the supposition that it possessed a yet-to-be-discovered 

 ventral armour. In the present state of our knowledge it is 

 scarcely possible to doubt that this was the case and that one day it 

 wiU be found. Only one or two specimens are yet known, and in 

 this respect the fossil resembles those species of Pteraspis and 

 Cyathaspis that are not yet known to have had a second plate. 

 Doubtless time and search will in all cases bring these to light. 



YII. The Lateeal Plates op Pal^aspis. 



Among the material in my possession, I from time to time observed 

 fragments which obviously could not belong to either of the two large 

 shields of the fish. As the number of these fragments increased a 

 general resemblance became apparent, and though I have not yet 

 obtained a single perfect specimen in place, yet the supposition that 

 occurred to me has proved well grounded, and scarcely a doubt remains 

 that the fragments represent lateral plates corresponding to the 

 ' cornua ' of Cyathaspis. I had not previously suspected the existence 

 of such plates. Their presence, however, brings Palceaspis yet more 

 closely into connexion with the rest of the Pteraspidian family as now 

 understood. At the same time the most careful examination fails to 

 remove a peculiarity in which this genus differs at present from all 

 other known Pteraspids, except perhaps Diplaspis. Instead of 

 extending along the whole side of the fish they were apparently 

 limited to the posterior half, and ended forward in points. They 

 thus fill what would be otherwise a narrow wedge-shaped opening 

 between the dorsal and ventral shields. I at first not unnaturally 

 regarded this hitherto unnoted and abnormal feature with much 

 suspicion, but the concurrent testimony of numerous specimens, 

 combined with a few lateral plates which gave more or less distinct 

 evidence of their position, have led me to regard it with confidence 

 as the actual arrangement and form of these plates. The signi- 

 ficance of this fact, then, is that the dorsal and ventral shields nearly 

 met or possibly were in actual contact (on this point see later) from 

 the front backward to about the middle, where the point of the 

 lateral plate came in ; and this gradually widened out to the hinder 

 margin, where it came to an end with the two shields. This 

 arrangement is shown in the lateral view of the type-specimen 

 (fig. 3), so far as it is perfect, and in fig. 4 (see p. 552). 



I have used the words " nearly met " above because there is some 

 reason for thinking that the two large plates were not in actual 

 contact, still less in sutural union, along their fore halves. The 



