OF FROGS OF THK GEXUS MEGALOPniiVS. 4(»1 



bet\ve9ii two subcutaneous lymph-sacs. At eacli oornevnear to the 

 angle of the jaw is the orifice into the mouth-cavity, through which 

 a probe was, of course, passed to show that the apei'ture did actually 

 communicate with the mouth-cavity. On opening the moutli 

 the orifices in question were very plainly visible posteriorly. It 

 will be noted that the gular sacs in this Frog, instead of being 

 separate sacs as in liana, are entirely confluent and form one 

 median ventral sac, which is shown in the accompanying text- 

 figure (text-fig. 128), as in Rhinoderma dai-vnni*. There is no 

 trace of a division in the middle line of the body between the 

 I'ight and left sacs ; they obviously form a continuous single 

 narrow sac lying just behind the subhyoideus muscle opening at 



Text-fiff. 128. 



/ 



/ 



Gular sac (S.) of Megaloplirijsfecs. 

 On each side the davk shaded orifice into tlie buccal cavitj' is seen. 



either extremity into the mouth-cavity. The walls of the gular 

 sac are rather extensive ; but, as has been already mentioned, they 

 are insei-ted in common on to the skin. This obviously would 

 hinder any very gi-eat dilatation of the sacs. Furthermore, they 

 could not extend in a dilated condition very far back along the 

 body, for the septum between the lymph-sacs in the pectoral region 

 would clearly prevent this. In Rhinoderma, on the other hand, 

 " the main portion of the sac hangs free " f. 



As to the musculature proper, the rectus ahdominis shows no 

 definite traces that were at all apparent to me of inscriptiones 

 tendinece, which seems to be rather an important difterence fi'om 

 it nearest allies. 



* Cf. Howes. "Notes on the Gular Brood-pouch of Ehinoderina darwini,'' P. Z. S. 

 1888, p. 231. 



f Howes, luc. cif. p. 234. 



