342 EDENTATA. 



from 48 to 56, and in 31. javanica it may even be as higli as 64. The extent to 

 wHcli these scales occur on the tail in the different species also affords materials 

 by which to separate the three forms. In 31. fentadactyla the average nvimber 

 is 14, in 31. aurita 16 to 20, and in 3£. javanica as many as 30. The Indian 

 ant-eater has also much longer fore claws than 3£. javanica, and in this respect it 

 is more closely allied to 31. aurita, but in all the species the hind claws are 

 shorter than those on the fore feet, but in 3£. javanica the difference that exists 

 between them is not at all well marked, and the claws are much less powerful than 

 in the two others. 3i. pentadactyla is distinguished from 31. javanica by its much 

 broader and less tapered tail, and in this respect it is closely related to 31. aurita, 

 the form to which it is most affined. In 3£. pentadactyla the tail stands to the 

 body and head in the proportion of 100 to 100, in 31. aurita of 71 to 100, and in 

 3£. javanica of 82 to 100. These proportions are gathered from measurements taken 

 from an authenticated skeleton of each species, as the limits of the tail are difficult 

 of determination in Museum skins. It was once thought that the species might be 

 recognised by differences in the size of the external ear, but I am doubtful that any 

 reliance can be placed on this character. M. pentadactyla is also at once recognised 

 from II. javanica by its broader and more conical head and less pointed muzzle, 

 in which respects 31. aurita more or less resembles it, whereas the head of 

 M. javanica is very narrow, and with a longer interval between the eye and the 

 nose than in the other species. 



To summarise the characters : 3£. pentadactyla is known by its heavy body ; 

 by its tail which is broad at the base, tapering gradually to a point, and equalling 

 the length of the head and the trunk ; by its large light olive-brown scales, of 

 which there are only from 11 to 13 longitudinal rows on the trunk and a mesial 

 line of 14 on the tail ; and by its powerful fore claws, the centre one of which 

 is somewhat more than twice as long as the corresponding claw of the hinder 

 extremity. 



31. aurita is distinguished from 31. pentadactyla by its less heavy body ; by 

 its rather shorter tail which has less basal breadth than 31. pentadactyla ; by its 

 smaller and darker brown, almost black scales in the adult, which are more numerous, 

 there being from 15 to 18 longitudinal rows on the trunk, 17 rows being the 

 normal number, and 16 to 20 caudal plates in the mesial line ; and by its strono* 

 fore claws, the middle one of which is not quite twice as long as the corresponding 

 claw on the hind foot. 



31. javanica is recognised by its body being longer and more attenuated than in 

 the two foregoing species ; by its naiTOwer and more tapered tail ; by its longer, 

 more foliaceous and darker olive-brown scales, of wlrich there are 19 longitudinal 

 rows on the trunk, and as many as 30 along the mesial line of the tail ; and by the 

 claws of the fore feet being not nearly so long as in M. aurita, and being but little 

 in excess of the claws of the hind feet. 



The variation in the number of longitudinal rows, more especially in the 

 case of if. aurita, appears to be due to one or other of two causes. When 17 rows 



