E, Ray Lanhester — On a new Seterostracom Fish- shield. 243 



designation "orbital" or not — are distinct circular perforations, 

 surrounded completely \>y calcareous test, the surface of which is 

 grooved by circular striations. In Bolaspis I have ascertained, by 

 cleaning out the matrix at the point in question, that the " orbital " 

 opening is simply a notch, not closed in on its marginal aspect by 

 the test. This condition is obviously a step from that of Scaphaspis, 

 devoid of any indication of the orbit, to that of Pteraspis, where the 

 orbits are completely involved in the shield. 



A further important difference between Pteraspis and Holaspis is 

 the absence from the shield of the latter of anything corresponding 

 to the large pair of cornual perforations, which I think we are 

 almost justified in calling branchial apertures. Though the marginal 

 extension of the shield was greater in Holaspis than in Scaphaspis, 

 and sufficed to involve a portion of the orbit, it yet did not — as it 

 did not in Scaphaspis — embrace the branchial aperture. 



In connexion with this matter, I may venture a suggestion as to 

 the series of six or seven pairs of ill-defined impressions seen in the 

 shield of Cyathaspis Banksii, as well as in Kunth's Cyathaspis integer, 

 and spoken of by him as "flache Hooker." It is not at all unlikely, 

 from their relation to other parts, that they indicate the position of 

 a series of branchial chambers. 



Further differences between Holaspis and Pteraspis are seen in th6 

 less pronounced development of the rostrum in the former, and in 

 its small spine, which is little more than an angular posterior pro- 

 longation. The minute characters of the surface-markings of the 

 shield in Holaspis are eminently characteristic, and require special 

 attention. Before speaking further of them, I may contrast thg 

 general form of shield of Scaphaspis and Holaspis. Whilst agreeing 

 in the apparently non-composite nature of its shield with Scaphaspis, 

 Holaspis differs from it in having the rostral region (always recog- 

 nizable in Scaphaspis, especially in Sc. Lloydii and rectus) produced 

 into a. broad well-marked rostrum : further, in the presence of orbital 

 notches, and lastly, in its broad posterior region, which is truncated 

 and characterized by lateral angles, in place of being gradually 

 rounded off. 



The test of the shield of Holaspis presents the characteristic 

 structure of the Heterostraci : its surface-striations would be undis- 

 tinguishable in small fragments from those of Pteraspis Crouchii. 

 The disposition of these striations on the surface of the shield is, 

 however, quite different from that of other genera, and is remark- 

 able for the large extent of rectilinear arrangement which it presents. 

 Among specimens borrowed from Herefordshire, for the purposes of 

 the Palffiont. Society's Monograph, I remember a large fragment 

 presenting rectilinear striation of this kind, which at the time I was 

 unable to refer to any known Heterostracous shield, and in Holaspis 

 that fragment finds its explanation. The disposition of the fine 

 grooves on the rostral region is such as to clearly mark out the limits 

 of what corresponds with the emarginated rostral constituent of the 

 shield of Pteraspis. Immediately above each orbital notch is an 

 ovally concentric arrangement of the grooves, and ground each 



