Yol. 52.] EATJNA OF THE KEISLEY LIMESTONE. 419 



shallowness of the axal furrows, appears to pass laterally into the 

 pleurae without any interruption. The furrows, however, separating 

 the axal rings are well marked, but weaker in the middle than at 

 the sides. 



The median terminal piece has been described above. 



The whole surface of the pygidium is ornamented with minute 

 granulations. 



Affinities. — The general appearance of the pygidium, particularly 

 with regard to the characters and direction of the first and second 

 pairs of pleurae, reminds us considerably of Ch. subulatus (Linnars- 

 son), 1 which occurs in the Trinudeus-schists of Western Gotland ; but 

 the bifurcated posterior extremity, owing to the projecting ends of 

 the third pair of pleurae, and the triangular instead of quadrate form 

 of the median piece representing the fourth segment, are conspicuous 

 differences. The point also at which the first and second pleurae 

 diverge from each other and become free spines is nearly equidistant 

 from the anterior and posterior ends of the pygidium in our Keisley 

 form, whereas in the Swedish species this point is on a level with 

 the posterior end of the pygidium according to Linnarsson's figure. 

 The pleurae of the third segment in our form are also distinctly 

 marked off from the axal portion by a transverse groove, whereas 

 in Ch. subulatus there is no such separation. 



The head-shields which I would tentatively ascribe to this species 

 have been mostly assigned to Ch. bimucronatus, but they do not 

 strictly agree with the typical form of that species nor with any of 

 the hitherto recognized varieties. The resemblance between the 

 pygidium of Ch. heisleyensis and Ch. subulatus has been noticed 

 above, and when we compare the head-shield of Linnarsson's species 

 with these of the so-called Ch. bimucronatus we cannot fail to 

 remark the close similarity. The chief differences between the 

 Swedish and Keisley forms are that in the latter the glabella is 

 broader in proportion to its length, is more parallel- sided, and the 

 eye is situated in a more forward position on the cheek. 



From Ch. bimucronatus the Keisley form differs in the glabella 

 being of a squarer shape and of greater width at the base, in 

 the lateral furrows being straighter and not directed so much back- 

 ward, and in the eye being placed a little in front of the second 

 lateral furrow instead of slightly behind it. 



Several hypostomes of a large Cheirurus have been found 

 associated with the above-described species, though not in actual 

 connexion or attachment. In point of size these hypostomes 

 would just suit the head- shields of Ch. keishi/ensis : they resemble 

 very closely the hypostorae of Ch. bimucronatus figured by Salter. 

 The central convex portion has its sides a trifle more parallel, and 

 the posterior end is rather broader and more obtuse, otherwise 

 there is no apparent difference between the two. 



Several disconnected thoracic segments belonging to large 

 individuals of Cheirurus have been found. It is probable that they 



1 Linnarsson, ' Ora Veetergotl. Cambr. och Silur. Aflagringar,' K. sv. Vet.- 

 AVad. Handl. toI. viii. (1869) no. 2, pi. i. fig. 5, p. 60. 



