F. R. Coirper Reed — On Dionide atra, Salter. 201 



and traces of a pair of small depressed triangular basal lobes at sides 

 of median spinose prolongation. Surface of glabella ornamented like 

 inner region of cheeks. 



Axial furrows united in front of glabella, witli small pit at front 

 end on each side. 



Thorax imperfectly known ; axis as wide as glabella, cylindrical ; 

 pleurae horizontally extended, straight, of uniform width, with fine 

 dia<ional furrow and obliquely truncate ends (?). 



Pygidium parabolic, flattened, nearly as long as wide ; axis conical, 

 elongated, about one-fifth the width of pygidium in front, tapering 

 to blunt end not quite reaching posterior margin of pygidium, com- 

 pletely annulated to tip by 20-24 well-defined rings, of which the 

 anterior ones have the interannular furrows trilobate, being arched 

 forward in the middle and at the sides. Lateral lobes horizontal, nearly 

 flat, composed of 1 5-1 7 well-marked straight radiating pleurae, widening 

 outwards but ending a short distance inside the margin (the last 5-6 

 pleurae arch slightly backwards and become subparallel) ; each pleura 

 marked by fine submedian slightly diagonal fuiTow reaching nearly to 

 extremity. No border present. 



Surface of thorax and pygidium very minutely punctate. 



Dimensions. mm. 



Length of head-shield . . . about lb 



Width of head-shield ... ,, IS'O 



Length of pygidium . . . . ,, 6'5 



Width of pygidium . . . . ,, 7'5 



Sori%on and Localities. Upper Arenig : Ty Obry, Portmadoc ( Salter's 

 types) ; Long Plantation Cutting, near Scolton, Haverfordwest 

 (Turnbull's specimens). 



Remarks. The pygidium of this trilobite as represented by 

 Mr. Turnbull's specimens is certainly identical with that described 

 and figured by Salter^ as Dionide atra from Ty Obry, Portmadoc; but 

 the two type-specimens are distorted, one being laterally compressed 

 and therefore unduly elongate in appearance (Salter, op. cit., fig. 9), 

 and the other being unnaturally broadened out transversely (Salter, 

 op. cit., fig. 9ff). In the latter specimen the inner surface is exposed 

 ^d a narrow doublure of uniform width is therefore visible. In the 

 elongated specimen (here figured) the axis shows 20-22 rings and 

 over 20 pleurae are present on the lateral lobes, while in the other 

 about 16-20 pleurae can be counted. Salter seems to have been 

 incorrect in stating that the pleurae are interlined ( = furrowed) for 

 only their first half and that the furrows bend back abruptly, but the 

 mistake arose from the poor preservation of the types. 



The head-shield and thorax of this species have been hitherto 

 unknown, and it is satisfactory to find them now in attachment to the 

 pygidium. The isolated head-shields are better preserved than 

 the more complete specimens, though those belonging to the two 

 individuals possessing much of the rest of the body show the characters 

 with quite sufficient distinctness. The structural characters of this 



^ Salter, Mem. Geol. Surv., vol. iii, p. 321, pi. xia, figs. 9, 9a. 



