248 



L. H. SNYDER AND \V. J. CROZIER. 



difference can be correlated with relative activity. New gills are 

 in all probability added at the posterior end of the ctenidial series. 

 Occasionally gill plumes are observed with divided tips (such as 

 have been figured by Pelseneer, 1920, p. 204). 



100 



Length of valve 4 in mm. 



FlG. I. 



It appears from Fig. i that the full number of ctenidia is estab- 

 lished before maximum growth is attained. It is unlikely that 

 there is a real decrease in the number of ctenidia with extreme age 

 (cf. Table I.). The number of gills is frequently mentioned as a 

 subsidiary character in descriptions of species, hence recognition 

 of the variation we record may be of taxonomic importance. 



3. This inquiry began with the thought that the total number of 

 gills might be influenced by habitat. The number of individuals 

 at our disposal is small, but sufficient for preliminary study of the 

 point. 



