0STE01,0GY. 41 



The metacarpals and phalanges are large and strong and in relative length 

 Wduld be arranged in the order 3, 4, 2, 5, 1. The middle three digits are sub- 

 e(juai, and have each three phalanges. The first and fifth digits are conspicu- 

 ously shorter; tlie former has but two phalanges, but the latter has three. 

 Tlu' claws are sliorter in S. pdrudoxus than in S. cubanus. A pair of small 

 sesamoids is present ventrally at the ])roxiinal end of tlie first jihalanx of each 

 digit. These as usual, form a groove for the great flexor tendim. A single 

 median scsanidid is also present at tlie ventral articulation of the two terminal 

 phalanges of each digit. In tlie pollex, however, it is very small. 



The femur (Plate 8, figs. 12, 13) is much like that of Gymnura in shape but 

 has a shorter shaft. Indeed, the shortness of the femur is remarkable. The 

 lesser trochanter is about as well developed as the greater. A deep fossa occurs 

 on the posterior side of the latter and is continuous with the broad intertrochan- 

 teric depression. A conspicuous sesamoid is i)resent on the ectal side, at the 

 posterior curve of the distal contlyle; it lies in tlie ectal tendon of the gastro- 

 cnemius. The large, subquadrate patella measures 7.5 nun. in length. The 

 greatest length from the proximal to the distal condyle of the femur is 47.5 mm. 



Dobson states that in Solenodon cubanus the tibia and fibula are distinct, 

 as in Centetes, Ericulus, Hemicentetes, and Tupaia although so closely approxi- 

 mated in the distal third that they might ankylose in aged individuals. Peters's 

 figure also shows the two bones distinct in this species. In the ailult S. puru- 

 doxiis, however, the two are united distally for 18 mm., or about one third their 

 length. The line of fusion is discernible, nevertheless, anil a deep groove is 

 formed between them, for the peroneus tendons. This point appears to be one of 

 some importance, for all previous writers have emphasized the separation of 

 til^ia and fibula in Solenodon as a character possessed in common with Centetes, 

 but not found elsewhere in the related families. It is not clear tliat the condi- 

 tion as thus described for S. cubanus is due to immaturity or individual variation, 

 but certainly in adult S. paradoxals, the tibia and fil)ula are distally ankylosed 

 as in Potamogale, except that in the latter the line of union is said to become 

 quite obliterated with age. It should be mentioned, however, that in the 

 skeleton of a young S. paradoxus examined, the tibia and fibula, though closely 

 approximated distally are ajiparently separate, and in the skeletons of two 

 fully grown though not old animals, these bones are Init imperfectly fused along 

 the line of contact. The antero-internal face of the tibia is rather flat |:>roximally, 

 but the crest at the ectal border oi this flattened area is scarcely developed, in 

 contrast with the condition in (lynuuira in which a prominent crest projects 



