14 STUDIES IN BIOLOGY, 



itself : this, which is known as the crystalline style, is to be regarded 

 as a secretion from the enteric epithelium." 



From these quotations the normal position of the crystalline 

 style would seem to be in a distinct caecum, not in the main line of 

 the enteric canal. Sabatier, however, makes no reference to the 

 position of this structure in other Lamellibranchs, and in no way 

 indicates that he thought the arrangement in M . edulis peculiar. 



M. latus and edulis (with M. magellanicus, which resembles it) 

 are so closely allied that we may expect to be able to trace readily 

 the homologous parts throughout all, and, taking any other organs 

 besides the alimentary canal, we can easily find the corresponding 

 parts in all three species. It were strange, then, if this should not 

 hold good also in the case of the alimentary canal. As far as 

 external appearance is concerned, the disposition of the alimentary 

 canal is very similar in M. latus and edulis, as is seen by comparing 

 Pigs. 13 and 14 (latus) with Figs. 15 and 16 (edulis). In the figure 

 of M. latus (Fig. 13) the pyloric caecum is drawn aside to show the 

 direct intestine, but in the natural position the caecum, is exactly 

 applied above the direct intestine in the middle line. 



Suppose, then, the parts thus arranged, and that the caecum, 

 instead of passing the direct intestine, should end just where the latter 

 ends, we should in that case have a figure exactly like that of M. edulis 

 (Fig. 15), as seen from above. M. latus exactly resembles M. edulis 

 in the position and relations of every other part of the alimentary 

 canal — mouth, gullet, stomach ("utricular stomach" of Sabatier), 

 cardiac caecum, recurrent intestine, anterior coil of the intestine, and 

 rectum ; it is likely, then, that there will be some correspondence in 

 the short tract between the stomach and the posterior adductor. 



I would suggest that the part called the tubular stomach by 

 Sabatier is not simply part of the stomach, but represents intestine 

 and caecum. Its structure as given above favours this supposition. 

 The upper part is already a semitube containing the style, and this 

 semitube tapers from before backwards exactly as does the pyloric 

 caecum in M. latus. The channel ventral to the crystalline style 

 would then represent the direct intestine of M. latus, and it is, in fact, 

 partly enclosed by two overhanging longitudinal ridges. A dia- 

 phragm below the crystalline style roofing over this channel would 

 form an arrangement exactly comparable to that in M. latus, making 

 allowance, of course, for the very abnormal development of the 

 pyloric caecum in that species. The smallness of the channel taken 

 to represent the direct intestine of M. latus may be accounted for by 

 the space round the crystalline style being pressed into service for the 



