CARADOCIAN CYSTTDEA FEOM GIRVAN. 385 



specimens ; in E 16028 the projection is ca. 2-5 mm. In three specimens (E 16026-7-8) 

 the convex face appears to be that towards the observer when the brachiole is on his 

 right hand ; the evidence of other specimens is uncertain. 



§ 97. Thecal Plates. — Diameters observed in the ordinary plates : — 



E 16026 . . . . . 1-3 mm. to 38 mm. 



E 16027 1-0 mm. to 3-9 mm. 



E 16028 1-3 mm. to 3-6 mm. 



From this it appears that the plates do not reach quite so large a size as do some in 

 D. Sedgivicld. But the difference that really exists does not appear from the measure- 

 ments : it is that here there are very few small plates, and all the plates are much of 

 a size, with a greatest diameter of ca. 2*5 mm. Thickness estimated from El 6030 as 

 not more than "25 mm. In most specimens any hollow moulds that may have been 

 left by the plates are too slight or obscure for measurement. Surface of plates smooth. 

 Stereom never preserved, and the nature of its reticulation obscured by the coarseness 

 of the matrix. 



§ 98. "Basals." There is at any rate on one face of the theca, a large plate. In 

 E16028 (PL I. fig. 4) it measures 4*1 mm. in greatest width, and 2"8 mm. in greatest 

 height ; it is surrounded by 9 (? 10) plates, having a width where they abut on it not 

 exceeding 1 mm., but widening as they radiate away from it, and attaining a height of 

 ca. 27 mm. The large plate had a very strong swollen rim at its base ; and in El 6028 

 this formed a rather sharp, slightly bifid projection, apparently on the inner side. In 

 E 16024 a strong rim is seen at the base of two adjoining large plates (PI. I. fig. 3). 

 In other respects there is no definiteness in the number or size of plates adjoining 

 the stem. 



§ 99. Adbrachial plates : a few are larger and stouter. In E 16027 one large plate, 

 3'7 mm. wide and 3 '2 mm. high, abuts laterally on two plates of rather less size, and 

 rests on 5 (or 6) smaller, somewhat elongate plates (see § 62). 



§ 100. The Brachial Appendage appears to have essentially the same structure 

 as in D. Sedgwicki (see §§ 63-66), but its study has been rendered difficult in all 

 specimens accessible to me by the absence of counterparts. In E 16028 (PL I. fig. 4), 

 the impression shows only a single series of half-rings, instead of the biserial arrange- 

 ment usual in the genus ; there are, however, a few wedge-shaped ossicles that thin 

 out towards one side or the other, and these suggest an incipient alternation of 

 ossicles. The study of such Crinoid genera as Platycrinus and Encrinus long ago 

 convinced us that the biserial arm in them and their allies was gradually evolved from 

 the uniserial. The brachioles of Blastoids and Cystids difi"er from the Crinoid brachium, 

 not merely in more fundamental features, but also in the fact that they are invariably 

 biserial and present no trace of an anterior uniserial stage. It would therefore be of 

 much interest if we could prove this slightly modified uniserial arrangement to be 

 characteristic of the earliest known species of Dendrocystis. 



