268 PROCEEDINGS OF THE NATIONAL MUSEUM. WOR, FSOIOV, 
first costals; the radials are, unfortunately, concealed. This twelve- 
armed condition is not due to the addition of an extra ray with a 
six-rayed rosette and six basal rays, as in the case of the six-rayed 
individuals of Zropiometra carinata, for two of the costal series 
are found between two of the small tubercles representing the ends 
of the basal rays, showing that these two series either both spring 
from a single radial or from a divided primitive radial. These 
costal series are also closely crowded, with their axes approximately 
at right angles, the-adjacent sides of the first costals projecting 
considerably from the centro-dorsal, the opposite sides mainly con- 
cealed beneath it. The arms borne by these two costal series and 
the arms on the four other costal series are normal in structure; 
but the outer arm of the left-hand one of the double costal series 
has the first brachial fused with the axillary of the adjacent costal 
series, the second brachial fused with the dorsal surface of the first 
brachial and lower part of the second 
brachial of the same series, with its pin- 
nule arising from a point over the sec- 
ond brachial of the normal arm, and its 
third and fourth brachials (syzygial 
pair) fused with the lateral edge of the 
second brachial of the normal arm; the 
remaining brachials of the two arms are 
free. 
The only other recorded case where 
Fig. 5.—HELIOMETRA MAXIMA; WITH two costal series arise from a single 
eae re sare = Oe" vo dial is recorded by Dr. P. H. Carpen- 
ter in Comaster alata (Actinometra 
pulchella) ; but here the radial bears two costal series which are 
smaller than the others, while in the specimen of /H7eliometra 
eschrichtii the two costal series are just as well developed as the 
normal ones. 
TI consider this specimen of Heliometra maxima to be a variation 
toward the condition found in Promachocrinus; and it seems to 
show that Promachocrinus may be derived from Heliometra by a 
simple division, or doubling, of the radials at an early stage of 
erowth, each resultant half of the primitive radial being of equal 
vegetative power. The basal rays of the adult Promachocrinus le 
under one of a pair of radials, instead of between the pairs, as 
would naturally be expected, this change in position possibly oc- 
curring through a predominance of pressure on one side of the anal 
plate as the latter is lifted out from between the radials. 
In Decametrocrinus, the ends of the basal rays are visible between 
the pairs of radials, and not under one of them, as in Promachocrinus. 
This suggests a fundamentally different condition in the early stages 
