224 Mr. F. A. Bather on British Fossil Crinoids : 
this in the earlier part of the paper (p. 209), that it is only 
necessary to point out that a Silurian species is here described, 
from specimens of mature growth, in which the axial canal 
is not separated from the ventral groove in the radial facet: 
in this point C. vallatus resembles Carboniferous species of 
Cyathocrinus. Further evidence, if such be needed, to show 
that the non-separation of the canal is merely a youthful 
character, and therefore also an archaic one, may be adduced 
from the young specimen at Jermyn Street (p. 211, fig. 5 a). 
Consequently it is not in itself a character very suitable for 
the discrimination of genera. 
The Covering-plates of the Arms.—It does not appear from 
the present paper, but it will be seen from the descriptions of 
the Gotland Cyathocrini that, although the number of these 
that goes to an ossicle is variable, yet there are limits to the 
variation, by the recognition of which we are often able to 
determine species when other means fail us. 
In describing these structures, Messrs. Wachsmuth and 
Springer have mentioned (Rev. I. 84, Proc. 1879, p. 307) that 
the groove “is provided with two rows of from two to five 
successive movable plates, alternately arranged on opposite 
sides.” It is, however, undoubtedly the case, as shown by 
Ph. XIE. fig. 10, that the row on either side may be only 
one plate deep. Tt is quite true that there are sometimes 
two plates in the row, a small narrow plate lying at the base 
of and alternating with each of the regular conical covering- 
plates (see Angelin, Iconogr. tab. xxvi. fig. 56). Occa- 
sionally too there oceur small, usually rather irregular plates, 
over the middle line, between. the two rows of regular 
covering-plates. This might make three or conceivably four 
rows on either side (see W.@ 5S. Rev. IIL. Proc. 1885, pl. iv. 
fig. 76). But in asserting that there were sometimes five 
rows it is possible that Messrs. Wachsmuth and Springer 
were misled by Angelin’s tab, xxvi. fig. 4, which represents 
the covering-plates of C. ramosus (wrongly called there C. 
longimanus) ; for these plates are marked by transverse lines 
that divide them into five parts, and produce the impression 
that each covering-plate is composed of five ossicles, which is 
not really the case. 
The Ventral Sac.—Messrs. Wachsmuth and Springer stated 
in 1879 (Rev. I. 84) that pores and slits had been observed 
in the ventral sac of Silurian species of Oyathocrinus. ‘This 
statement has never been withdrawn by them, though in their 
recent paper on the Perisomic Plates * they seem to imply 
* Proc. Acad, Nat. Sci. Philadelphia (1890), Part III. See p. 360, 
February 1891. 
