394 MR. A. H. G. DORAN ON THE MORPHOLOGY 



except Arctictis, and forms a very obtuse angle with the neck ; the processus brevis is 

 generally well marked. The short, stout, conical processus muscularis does not spring 

 from the neck, but from the inner edge of the manubrium, a perceptible distance beyond 

 its root, approaching a characteristic of some rodents. The lamina is very little deve- 

 loped, indeed is not perceptible until close examination. 



The incus (PI. LVIII. fig. 47) is rather larger than in the Civets ; the stapes is large 

 (PI. LVIII. fig. 61), and resembles that of Genetta and Arctictis. 



The species do not differ much in their ossicula. The head of the malleus in Herpestes 

 ichneumon and in H. paludosus is particularly large. In the manubrium of the latter, Prof. 

 Plower has detected a minute foramen which perforates it close to the root of the pro- 

 cessus muscularis. In the above species and in S. nepalensis, where the head is smaller, 

 the processus brevis is well marked ; but it is much blunter in H. griseus and JEL. pulve- 

 rulentus. In the latter (PI. LVIII. fig. 38) the processus muscularis is a more con- 

 siderable distance down the inner edge of the manubrium than in the malleus of any 

 other ichneumon ; its incus is proportionally larger. 



In Suricata zenik (PL LIX. fig. 3) the ichneumon type of malleus is maintained. The 

 head is large, the neck extremely short. The manubrium leaves the neck at a wider 

 angle than in Herpestes, and is distinctly shorter, broader at the base, and more spatulate 

 at the tip. The processus muscularis is situated on the inner side of the manubrium, a 

 perceptible distance from the neck; and the lamina is very little developed. 



In the Cryptoprocta the malleus closely resembles that of Viverra ; the head is a little 

 larger, the articular surface similar ; the manubrium is even more slender ; the processus 

 muscularis is almost as short as in Paradoxurus. The incus (PL LVIII. fig. 44) is a 

 larger bone than in the Civets ; and the stapes (PL LVIII. fig. 59) has thicker crura than 

 in Viverra. 



Thus Cryptoprocta ferox is incontestably more allied, as far as its ossicula are con- 

 cerned, to the Viverridas than to any other family ; and the latter are clearly more 

 related to the Cats by the type of their ear-bones, as Hycena and Proteles are related 

 to the Dogs. 



In the Canidse the ossicula have far more marked distinctive features than they bear 

 in the Viverrida?. 



The malleus (PL LVIII. fig. 33, and PL LIX. fig. 5) often rivals that of the Hyaena 

 in stoutness of build. The head cannot be said to be remarkably developed; the articular 

 surface is narrow, and deeply cut ; both facets are concave ; the groove dividing them 

 never bifurcates, as in well-marked viverrine mallei ; but the facets join in a concave area 

 external to the point where the groove fades, as in Proteles. The neck is only mode- 

 rately curved ; the manubrium, the most distinctive feature in the canine malleus, is 

 long, rather thin at the root, and very strongly curved, with the concavity looking 

 forwards. The outer aspect is a distinct surface, not only sharply bordered from the 

 sides, as in all Carnivora, but likewise very broad throughout, though rather broader 

 towards the extremity. Thus the Hysenidse, as has been already remarked, resemble the 

 Dogs in their manubrium ; but Proteles is still more canine. On the other hand, the 

 Canidae have a long sickle-shaped processus muscularis projecting from the inner side of 



