INSTITUTE OF SCIENCE, PHILADELPHIA. 63 



of compression. Shells without lirje should, by the hypothesis, in general 

 have a mantle-margin without papillae or with very small ones, a short ad- 

 ductor muscle, or a wide aperture, or all these characters combined. 



Those with extraordinary denticulations of the aperture should be capable 

 (as in Distortrix) of widely extending the mantle, or it should have its edge 

 peculiarly modified. In many cases, such as the inoperculate land shells, the 

 denticles would serve as a protection and be laid hold of and developed by 

 natural selection in various ways. Finally, after a physiological habit of 

 forming a certain set of lirje (or even plaits) has been acquired, we might ex- 

 pect them to be retained for some time, even if the original conditions which 

 induced their initiation should to some extent be changed or partially lost in 

 the course of the general evolution of the organic type. 



Obliquity in the Plaits. — The older conchologists were in the habit of plac- 

 ing much stress on the aspect of the columellar plaits in matters of classifica- 

 tion — as, for instance, whether the plaits are obliquely coiled or more nearly 

 at right angles to the axis of the pillar. In this many modern naturalists have 

 copied the old diagnoses without discrimination, not detecting that, with the 

 progress of investigation, the facts now available make the old diagnoses in- 

 applicable. The existing text-books generally emphasize such statements as 

 this : that the VolutidcB are separated from the Mitras by the more horizontal 

 plaits of the latter, and by the anterior plaits in Valuta being larger, while in 

 Mitra the posterior plaits are more conspicuous. But a complete series of 

 VoliitidcB will show that almost every variety of horizontality or obliquity may 

 be found, that the plaits are sometimes unequal and sometimes equal in size, 

 and that the more prominent of them may be situated in either part of the series. 



It is in order, therefore, to consider the circumstances which are associa- 

 ted with the different types of plication in this family, and to what dynamic 

 conditions these differences may be with probability ascribed. 



If one winds a cord about a cylinder, as an ordinary cylindrical lead-pen- 

 cil, it will be observed that in proportion as the number of turns diminishes in 

 the same length of pencil, the obliquity of the plaits increases. The same 

 thing is observable in the " lay " of a rope, and if a given length of rope be 

 additionally twisted it will at the same time be shortened. 



The obliquity of an applied plait will also decrease in the same length in 

 accordance with the increase in the diameter of the cylinder to which it is ap- 

 plied. The principle at bottom is the same in this as in the preceding ex- 

 ample, for with the same axial length of cylinder the actual length of applied 

 plait will be greater the greater the diameter of the cylinder, and it is really 

 only another example of the proposition that in the same axial length the 

 greater the length of the cord coiled the less oblique will be its direction. 



These principles are amply illustrated among the animals under con- 

 sideration. 



Take the group called Caricella by Conrad, which is the direct ancestral 



