136 bulletin: museum of comparative zoology. 



in other zalambdodonts — Centetes, Potamogale, Microgale, etc. — 

 the cusp-deAelopment seems to result in the formation of a single V 

 instead of the more primitive W-pattern. The generally accepted 

 view is that the main cusp represents a fused paracone and metacone 

 in which the latter is the obliterated element. In Solenodon two small 

 internal cusps are interpreted as protocone and hypocone, the large 

 median cone as the fused paracone and metacone. Another view 

 may be put forward, however; namely, that the main cusp in the 

 Solenodon molar is the metacone, and that the paracone is represented 

 by a small style-like eminence anteriorly at the outer rim. At all 

 events it is frequent in the Chiroptera to find the paracone much 

 smaller than the metacone, while protocone and hypocone are small, 

 at the inner edge of the tooth. The point of interest is that the upper 

 molars of Nesophontes offer a similar intermediate condition between 

 the fully developed W-pattern of cones and the V-pattern of Solenodon 

 and perhaps other zalambdodonts. For although Anthony states 

 that the molars of Nesophontes have the "V-shaped metacone only, 

 instead of both metacone and paracone," it seems clear that this 

 appearance must have been due to wear in the specimens he studied. 

 Most of our series of crania of N. micnis from Cuba are in a similar 

 condition, but one nearly perfect specimen has the teeth unworn. 

 These show that the first and secontl molars have a large protocone 

 forming the main cusp of each tooth, with a low cingulum-like ridge 

 representing the hypocone. A small paracone is clearly developed, 

 about one fourth the area of the metacone. With wear, the former 

 disappears; yet its outlines seem traceable even in Anthony's figure 

 (Fig. 3) of the palatal view of N . cdithac. A further reduction in the 

 size of the paracone and protocone, with corresponding enlargement 

 of the metacone, would easily give rise to the type of molar seen in 

 Solenodon. It is not so obvious that this method of derivation would 

 appl\- equally to the Old World zalambdodonts, but it seems fairly 

 apparent in these two genera of the New World. If this view be 

 accepted, Nesophontes may represent a stage in molar-development 

 ancestral to that of Solenodon, in which the prominent features of the 

 skull and limbs are similar, but the teeth are much more primiti^•e, a 

 stage perhaps in some degree intermediate between the zalambdodonts 

 and the soricoids, though distinctly nearer the former. 



The humerus is almost a miniature of that of Solenodon, but more 

 slender, with globose head, strongly marked bicipital groove and 

 deltoid ridge, the last so elevated as to accentuate considerably the 

 compressed form of the shaft. At the distal end, the humerus is 



