1908.] OF THE BATRACHIAN GENUS HEMISUS. 933 



distinguishing features of Hemisus, besides certain osteological 

 and external peculiarities which I do not deal with here. Of the 

 former there are more that are peculiar to Hemisus than of those 

 which ally it to its allies. Hemisus is peculiar, so far as is at 

 present known, in Nos. (1), (2), (3), (6), (7), (11), (13), (17)— that 

 is, in nearly one-half of those which I have selected. 



Hemisus agrees with both of its allies, Breviceps and Rhino- 

 derma, in (5) — at any rate, as to the fact that there are variovis 

 additional muscles, not present in liana, upon the floor of the 

 mouth ; there is, however, no detailed agreement between the 

 three genera in the disposition of these muscles. In (10), (12) 

 Hemisus agrees with Rhinoclerma, and differs from Breviceps 

 in (16). Hemisus agrees with Breviceps and differs from Rhino- 

 derma in (4), (8), (14). I am not quite certain as to the re- 

 maining featui-es of the anatomy, which I have made use of as 

 indications of closer or more remote affinity. These facts, and 

 indeed others which will be found in the foregoing pages, do 

 not, as it seems to me, permit of a very decisive placing of 

 Hemisus with regard to the two remaining genera of Engysto- 

 matid Frogs whose anatomy is known. The pai'ticular likenesses 

 which Hemisus shows to BrevicejDS, as opposed to Rhinoderma, 

 may be increased by the addition of the fact of the partial 

 inclusion of the limbs within the area of the trunk, and by the 

 division of the rectus abdominis muscle by only a single inscriptio 

 tendinea. But Breviceps remains, after all, an extremely specialized 

 type in many ways. General reflections upon the arrangement 

 of these Frogs will, in fact, be better deferred until more 

 anatomical facts have been collected. 



§ Resume of principal neio Facts. 



It may be convenient to extract from the foregoing account of 

 Hemisus, and of Xenop)u,s and Rana, the following principal new 

 facts which I have been able to ascertain : — 



(1) Hemisus is characterized by the existence of three pairs of 

 large-lobed fat-bodies, of which one pair correspond in 

 position to the thymus in other Frogs, the second lie in a 

 cavity (? a lymph-sac) behind the shoidder-girdle, and the 

 third pair are contained in a sac partly overlapping the 

 thigh, which is to be compared with the saccus iliacus of 

 Rana. The prerenal fat-bodies are also very large. 



(2) Xeuopus has a similar pair of fat-bodies in the repre- 

 sentative of the saccus iliacus, the fatty tissue, however, 

 straying further forward on to the back. 



(3) The saccus iliacus in Hemisus and Xenopus is divided by 

 trabeculfe in the interstices of which lies the fat-body; 

 the commencement of such a division of the saccus iliacus 

 is seen in Rana guppyi. It is possible to compare these 

 structures with lymph-glands. 



