from, the Neotropical Region. 313 



side of the metatarsal in some species — notably in E. trinitatis^ 

 All. & Chap. — little marked. 



Skull, as compared with those of the specimens assigned to 

 E. semispinosus, distinguished by the following characters : — 

 The nasals run evenly backwards to a sharp median point, 

 instead of being more or less bluntly truncated behind. The 

 ankylosed parieto-interparietal suture runs nearly squarely 

 across, instead of being evenly convex forwards, and from its 

 outer corners little ridges run forwards to meet those that run 

 back from the supraorbital ledges, while in the allied species 

 the posterior ridges are at a higher level and do not coalesce 

 with the anterior. Zygomata much broader both in the 

 ascending and horizontal portions. Anterior palatine foramina 

 narrower and less widely open. Posterior nares more widely 

 open. Pterygoid processes broad and spatulate, over 2 millim. 

 in their greatest diameter, as compared with about 1 millim. 

 in E. semispinosus. 



Dimensions of the type (an adult male) : — 



Hind foot, without claw 49 millim., with claw 53 *. 



Skull : basal length (c.) 47'5j basilar length 41 j greatest 

 breadth 26*6 ; nasals 21-3x6; interorbital breadth 12*5; 

 breadth of interparietal 12"5; palate length from henselion 20 ; 

 diastema 12'5; anterior palatine foramina 5x2'5; length of 

 upper molar series 9*2. Lower jaw : back of condylar process 

 to incisor-tip 34. 



Hah. San Emilio, south end of Lake Nicaragua. 



Coll. W. R. Richardson. 



Type obtained March 27, 1896. 



The above differences, slight as they seem, are perfectly 

 constant through series of seven of the Colombian and two 

 of the Nicaraguan form, and it is therefore evident that the 

 two animals cannot be regarded as both belonging to E. semi- 

 spinosus. E. trinitatis, again, though belonging to the same 

 group, is evidently quite distinct from either of the others. 



Marmosa fuscata, sp. u. 

 Size rather smaller than in M. incana, Lund f. Ears 



trap which one has to be constantly on oue's guard against in describing 

 from, modern skins. Another slip in the same account is that the molar 

 series is said on p. 224 to be 10 millim. long and is given as 8'0 in all the 

 specimens measured on p. 226. The latter is evidently correct. 



* The skin is too badly made to make other external measurements 

 worth taking ; but Mr. True (J. c.) has given accurate measurements of 

 several spirit-specimens of what is no doubt the same species. 



t Dideljihys grisea of Cat. Mus. B. M. p. 349, but not of Desmarest. 

 For the real M. grisea an earlier name has since turned up in Uidelphys 



