594 
first whorl, slight contact furrow of an ordinary transitional form, but 
otherwise the nepionic stage resembles the adults of Aphetoceras in 
its section and position of siphuncle and sutures. 
The gap between Pycnoceras and the next member of this series, 
Tarphyceras, is wide and one or more genera are needed to fill up 
the interval. 
In all of the genera mentioned above, except Tarphyceras, there 
is no dorsal furrow, the zone of impression is produced by contact, 
and the umbilical perforations are large. 
In Tarphyceras, however, although in form, sutures and position 
of siphuncle the genus is closely allied to Aphetoceras, the young 
are altogether distinct. 
As depicted on Pl. iv, the young have very small umbilical perfo- 
rations, the whorls broaden out by growth rapidly, and after a short, 
straight or only slightly curved apical part is built in the ana- and 
-metanepionic substages, the broadening volution makes a sudden 
and very abrupt gyroceran bend towards the apex. This is very 
sudden and the umbilical perforation is flat or comma-shaped. 
It might of course be shown, if other intermediate shells were 
found, that the mechanical effects of this sudden bending did not 
produce the dorsal furrow, but that this is an adequate mechanical 
cause can reasonably be claimed by those who oppose the view that 
it is due to heredity. 
It has already been shown that the outer side or venter tends to 
grow faster than the inner, and if this reaches a point in its ratio of 
growth that far exceeds that of the inner side, it is obvious that it must 
act upon that side as a force that bends or tends to make it more 
arcuate in proportion to this excess of growth or rapidity of increase. 
The outer side being free would be apt to retain its genetic 
form, and the inner side or dorsum would be greatly influenced or 
moulded by the pressure to which it was subjected. Thus it can be 
assumed that in case of a sudden bending, as in Tarphyceras, the 
venter would maintain its rounded outline and forcing the dorsum 
inward as it grew would tend to make it assume the arcuate form or 
bend inwards in a crease or dorsal furrow in the paranepionic volu- 
tion conforming more or less with the shape of the dorsum of the 
metanepionic volution. 
There are some reasons why this explanation is not wholly satis- 
factory. In the first place, if this be the case, why did not the 
whorl of the paranepionic completely close the umbilical perfora- 
