586 
riably towards the dorsum, but that this is the general tendency of 
arcuate forms is obvious. 
When it comes to the evolution of coiled forms the problem is 
different. Among these last, including also the loosely coiled or 
gyroceran, there is so far as known no exception to the rule that all 
such shells are bent dorsally from the earliest substages of the 
conch. 
I have assumed in other papers that coiling among Gasteropoda 
could be accounted for by the unequal growth caused by the weight 
of the shell when carried above the foot and the facts appear to 
justify this conclusion in so far as that class is concerned. ‘The 
presence of the hyponome on the ventral side in Cephalopoda 
would of itself account for the tipping of the shell towards the 
opposite or dorsal side both when the animal was crawling and 
swimming. ‘This would leave the ventral edge of the mantle free 
to deposit calcareous matter undisturbed by pressure, whereas the 
dorsal edge would be more subject to disturbance and to shocks 
from compression which might interfere with the work of excretion. 
It is reasonable to suggest such a mechanical explanation both for 
the general tendency to bending and coiling and also for the dorsal 
direction. 
If it were possible to account for the exceptions observed, as in 
the tendency of Phragmoceras towards the venter, by means of 
exceptional habits or structures, this suggestion would have more 
force, but unfortunately this cannot be done, at least at present. It 
is obvious, however, that there is some directive cause which acts 
upon every genetic series in greater or less proportion, giving to 
each evolving series the same tendency to produce in succession 
the straight, arcuate,* and then the coiled forms in different de- 
grees of intensity and that most of these have coiled in a dorsal 
direction away from the hyponome or organ of locomotion. 
The position of the siphuncle with reference to the mode of 
coiling need not be discussed, since it obviously has no general 
relations, except that it is always, except in turbinate forms, Tro- 
choceras, Turrilites, etc., and in abnormal forms, like some species 
* Even in large and some small species and specimens of Baculites there is an arcuate 
tendency. D’Orbigny figures this in B. incurvatus, Terr. Crétac., Pl. exxxix, and Mr. 
Stanton has put together a very large specimen of Baculites in Nat. Mus. having a curye 
like Cyrtoceras. This is some five feet long, straight or nearly so in the younger part 
and arcuate in the older stage. 
This specimen is from the Ripley Formation, Texas. 
