Occasional Papers of the Museum of Zoology 57 



the following longer account : The length of the shell is one- 

 half, the diameter is one-third, and the axis is at one-fourth 

 ■ of the breadth, or largest dimension of the shell. 



"In longitudinal shells this is reversed, the length being the 

 longest dimension becomes the size of comparison." 



Vanatta ('15, p. 549) seems to have overlooked this state- 

 ;ment, and consequently his explanation of the meaning of 

 Rafinesque's fractions is obviously incorrect. 



In course of the investigations made in the preparation of 

 this paper we have found Rafinesque's comparative measure- 

 ments in most cases very exact and are of the opinion that 

 they are to be relied upon as a means for identifying many 

 of his species. The discrepancy pointed out by Walker ('18, 

 p. 5) is an important one and a serious objection to the approx- 

 imation of riuwsus to this species. 



But this much is sure, that Rafinesque describes his shell as 

 elliptic, which cculatus is not, and that he describes it as broader 

 behind, which again does not fit. It is also correct that the 

 '"rimose'' character of the shell in ccelatus is not restricted to 

 the posterior part, while the main part of the surface is smooth, 

 as stated in the description of rimosus, but that it covers nearly 

 three-fourths of the entire shell. 



The term "rimose" might also be applied, according to Wal- 

 ker, to Medionidiis conradicus (Lea), and the rest of the 

 •description would not speak against this. 



"In discussing U. rimosus Raf. Frierson claims that by 

 'Broader . . . behind' Rafinesque means what we would 

 term longer behind ; but even so, the dimensions do not fit. 

 Rafinesque's shell measured, L. 15^, H. i, D. ^ inch. A speci- 

 men measures, L. i^, H. i}i, D. ^ inch. This is a rather 

 compressed and low example, but more than double the diam- 

 •eter assigned by Rafinesque. It may be claimed that Rafin- 



