1921 ] N. Annandai.e & R. B. S. Sewell : Vivipara. 265 



slight hypertrophy of the calciferous gland in a position on the 

 mantle corresponding with them on the shell. This is indicated 

 by the facts that even in smooth shells of V. bengalensis the 

 dark spiral bands are slightly thickened and that at the end of 

 a growth-period the calciferous glands are often a little larger 

 immediately above the primary marginal processes than at 

 other points on the periphery. It is clear, however, that some of 

 the matter which occupies their base is nacreous, and we know 

 that on the internal surface of the shell nacre can be deposited by 

 almost any part of the mantle after the external ornamentation is 

 complete. 



The third, most highly sculptured type of shell is the most 

 interesting of the three, not only because of its peculiar facies but 

 also because it has appeared and become dominant among the 

 Vivipandae ' on different occasions and in different places and 

 different geological epochs. The test sculpture, even in this 

 type of shell, corresponds closely in fundamental pattern with 

 the primitive periostracal sculpture of the embryo of V. bengalensis. 

 that is to say that it consists essentially of spiral ridges 

 bearing prominences and that these ridges have fundament- 

 ally a definite number and position on the shell exactly similar 

 to that of the three rows of chaetae and the secondary' ridges 

 of the embryonic periostracum, and that the most prominent 

 ridge corresponds with the peripheral row of chaetae. It follows 

 that the interrupted ridges of the test in this type of Viviparid 

 shell are correlated at least in position with the marginal processes 

 of the mantle, but the connections between the structures on the 

 shell and those on the soft part are certainly not so close as in the 

 periostracal sculpture and cannot be stated with the same pre- 

 cision. Here again, however, we know that the test sculpture is 

 not correlated as in V. oxytropis with any hypertrophy of the 

 marginal processes of the mantle, which are small in both Taia and 

 Margarya^ and also that the processes show no essential difference 

 of structure in individual shells of the former genus in which 

 the sculpture is less and more highh^ developed. 



If the mode of construction of the projecti<ms on the peri- 

 pheral and other ridges of the more highly developed shells 

 of the genera Taia and Margarya can be explained, that of the 

 remainder of the ridges is a simple matter. They cannot have 

 been formed, so to speak, in the air (or rather in the water) but 

 must have been built up in continuity with the edge of the 

 lip. In the fossil Rividarioides they may be nearly half as long 

 as the diameter of the whole shell. Their form suggests that they 

 must have been moulded by some comparatively broad projection 

 of the mantle edge. In contracted specimens of Taia and Mar- 

 garya no trace of any such structure can be detected, and the 

 extreme margin differs little from that of V. dissimilis, which has 

 a very smooth shell. ^ It is unfortunate that I made no observa- 



' Annandale, Rec. Geol. Siirv. Ind. I,, p. 209, pi. xxxi, figs. 8-1 1 (1919). 



