198 GENESIS OF THE ARIETID. 
parallel on the latter part of the second whorl; the dorsum also broadens, and 
the umbilical shoulders become more abrupt. On the third volution the sides are 
parallel, the dorsum and abdomen of the same breadth, and the latter elevated 
into an obtuse ridge. ‘This ridge on the fourth whorl becomes a true keel, un- 
accompanied by channels, though there are traces of their formation. The in- 
crease of the abdomino-dorsal diameter of the whorl after this period speedily 
elevates the abdomen, and proportionally decreases the transverse diameter of 
the shell on the latter part of the fourth volution, destroying even the faint 
traces of the channels mentioned above. On the fifth volution the sides become 
slightly convergent, and this convergence increases very slowly on the sixth 
volution. 
Thick and widely separated folds appear on the early part of the third whorl, 
but rapidly grow into true pile. The genicule may become tuberculated on the 
fifth volution, or they may remain permanently without tubercles; there is, how- 
ever, great variation with regard to their prominence and number. 
The duration of the different stages of growth also varies; in some specimens 
the abdomen remains flattened through the keel-forming stage until the first 
quarter of the fifth volution is reached. In one specimen the pile are reversed 
in position, bending posteriorly, but subsequently begin slowly to change to a 
natural position on the fifth volution. A number of the specimens from Semur 
have the pile but very slightly or not at all developed. None of them seem to 
reach beyond four and a half volutions, and their small size may possibly be due 
to the same cause as the absence of the pile. 
On the early part of the fourth volution the abdominal lobe is considerably 
longer than the superior lateral lobes, which in turn are shorter than the infe- 
rior laterals. The inferior lateral saddles are slightly deeper than the superior 
laterals. On the latter part of the same volution the abdominal lobe has in- 
creased in length, and the superior lateral saddles are shallower; the inferior 
lateral saddles become deeper proportionally, and the lateral lobes are nearly 
equal in length, the inferior laterals still remaining, however, slightly the long- 
est. The marginal lobes are very minute, and remain so, the sutures having a 
comparatively even outline for that reason. In some specimens the lengthen- 
ing of the abdominal lobe continues until the superior lateral saddles are almost 
obliterated ; in others, the lobes and saddles may remain with about the same 
proportions as in the stage of development last described. 
There are two specimens in the Stuttgardt Museum, from the Geometricus 
bed, but they do not show the young, or give any better clue to the true devel- 
opment. There are also three in the Stuttgardt Museum, from the Pentacrinus 
bed near Krumenacher, which appear to be the young of this species, or of 
Scipionis. One of them, with the shell present, shows all the peculiar marks of 
Agus. striaries, while another, a cast, is so broad on the abdomen, and the fold- 
like pile are so prominent, that it looks remotely like the young of Cor. Sauze- 
anum. The three are identified as young of Cor. Sauzeanum, but the young of 
this last is very distinct in the sutures and other characteristics. A specimen in 
the collection at Semur, from the same horizon, under the name of Amm. lavi- 
