GASTEROPODS. 171 



single one of many to be found among the mollusca. It is very 

 significant in its bearing upon the true basis of species, and indi- 

 cates plainly, that in attempting to separate specimens specifi- 

 cally, too much stress should not be placed upon individual 

 characters. 



Other Causes of Variation. In connection with variation of 

 species, it is of great interest to note the apparent effect of 

 gravitation in altering the form of some gasteropod shells. 

 This phase may be more satisfactorily considered in Capulus 

 equilateralis and I. pabulocrinus than in most other species, be- 

 cause when attached to the vaults of crinoids, the station of 

 each individual is definitely known. As stated already, the 

 first of these forms generally rests on flat-vaulted crinoids, 

 while the second commonly adheres laterally to such echino- 

 derms as Platycrinus hemisphericus Capulus equilateralis when 

 occupying the same position is pendant, the apex of. the shell 

 being directed downward instead of in the opposite direction, 

 as when resting on the ventral surface of such forms as Gil- 

 bertsocrinus. The shell thus pendant exhibits a decided ten- 

 dency to straighten, or uncoil, consequently becoming longer, 

 the apex freeing itself completely from the body-whorl. In 

 comparison, therefore, with a representative example of C. 

 equilateralis, those shells resting on flat crinoidal vaults are 

 very much depressed, the aperture proportionately broader 

 and the spire more closely coiled. Those individuals attached 

 laterally to crinoids have a tendency to become more conical, 

 the aperture being relatively smaller, while the spire is entirely 

 free from the last volution, and the apex often extends to a con- 

 siderable distance beyond the posterior margin of the aperture. 



On the other hand, L pabulocrinus is commonly a more or 

 less elongate conic shell. When attached to Platycrinus it 

 often assumes a very different aspect. As growth proceeds 

 the posterior side becomes relatively shorter, the apex slightly 

 curved backward, and not unfrequently there is a marked 

 tendency toward a strongly arcuate form. 



The Kinderhook forms of the genus are, on the whole, ex- 

 tremely unsatisfactory for systematic determination, since the 



