6 
The departure in outline from D. barrandei is very slight, but there 
were other changes. 
The marginals became more pronounced, more regular, and fewer, 
thus constituting a more obvious frame. 
A slight decrease in number of somatic plates is to be observed in D. 
scotica, but along the American line this decrease has proceeded more 
rapidly, reaching what must be almost an extreme on the reverse side of 
D. sagittaria . 
With the enlargement of the plates is naturally connected a develop- 
ment of axial folding and ridging. Traces of similar folding have already 
been described in D. sedgwicki 1 . 
The evolution of the stem also has proceeded along a line somewhat 
different from that of the European forms. The median and distal regions 
of the stem consist of alternating dimeres, which become shorter in the 
median region and are occasionally accompanied by intercalated plates. 2 
On the whole these parts of the stem are nearer the stage of the Middle 
Ordovician D. sedgwicki than of the more flattened Upper Ordovician D. 
scotica . The proximal region of the stem, on the other hand, has progressed 
further in the direction of flanged rings (7 or 8 in all) than in any other 
completely known Dendrocystis. The proximal ossicle of the median 
series fits into the lowest of these rings, and reminds one of the conical 
reducing piece in Cothurnocystis, Ceratocystis , and other genera (the stylo- 
conus of Jaekel). So far as the stem is concerned, the intermediate stage 
between D. sagittaria and D. barrandei is provided by the Trenton Limestone 
fossil which E. Billings (1859) named Syringocrinus paradoxicus, and 
which I referred to Dendrocystis in 1913 (par. 155). 
Some new’ morphological features are described by Messrs, Thomas 
and Ladd. To our knowledge of the brachiole itself, they unfortunately do 
not add; but they show that, with the reduction of the somatic plates, three 
of them have become more definitely associated with the brachiole than 
have any of the adbrachial plates in the European species. 3 One of 
these three might also be compared to the radial of a crinoid, and it would 
be interesting to see if it shows any trace of an articular facet for the 
brachiole. What is an important observation is that these plates bound a 
groove opening into the theca. The brachiole appears to be attached to 
the lower or proximal end of this groove. This is additional confirmation, 
if such were required, of my interpretation of this structure as a brachiole 
(1899, 1900, and 1913, par. Nos. 71, 72). 
Another of these three plates — that on the left when the theca is viewed 
on the obverse face — is “thick and the adapical [i.e. adoral] part of its 
surface is elevated into a pointed cone directed outward and upward; in 
the apex of the cone is a small depression, evidently a pore, thought to be 
the gonopore” (page 8). “On the flanks of the cone are a 
number of tiny pustules which have the appearance of being perforated; 
their purpose is unknown” (page 10). This is veiy suggestive of a madre- 
porite bearing a gonopore. Hitherto no such openings have been observed 
in any Heterostelea other than the pores detected by Jaekel (1919) on 
»Batber f 1913, par. No. 46. 
*Cf. Bather, 1613, par. Nos. 82, 83. 
*Bather, 1913, par. No. 01. 
