I92I.] 



Manipur Molluscs. 



561 



series of this variety I find shells of the form here called lacvis 

 mixed with others to which the definition he adopted from Benson 

 applies more exactly, the' essential difference being that in the 

 obsolescent ribs of semilaevigata both extremities are produced into 

 tubercles. Highly sculptured shells of this variety often approach 

 Nevill's var. bijiodulifera. 



Var. subspinata, nov. : pi. vi, fig. 6. This form is interesting 

 as being precisely intermediate between the var. binodulifera, 

 which occurs in abundance in the river at Dimapur just north 

 of the Naga Hills, and God win- Austen's M elanoides spinata from 

 northern Assam. The two rows of spines are nearer together 

 than in binodulijcra and the spines are more produced, but they 

 are not foliaceous as in spinata. 



Fig. 10. — Living animal of .Acrostoma variabilis \ar. subspinata, 

 nov., from Dimapur, Assam. 



These three varieties are, as I have said, by no means con- 

 stant. The two last occur together and in most individuals can 

 be readily distinguished even before the shell has attained a third 

 of its full dimensions, the longitudinal ribs being well developed 

 in this stage as strong keels on all but the first four or five whorls 

 in subspinata. while the whole shell is smooth in lacvis. Inter- 

 mediate individuals occur, however, not uncommonly. They are 

 much commoner among adult than among young shells. The 

 vars. lacvis and semilaevigata also occur together, but I have not 

 found both semilaevigata and subspinata in any series examined. 

 This would suegest that the smooth type of shell has been derived 

 independently along two lines, by the suppression of the lower 

 part of the ribs in one line and l)y that of the middle region in 

 another. 



