366 MR. R. I. POCOCK ON THE 



From the accounts above given, it is clear that the feet of 

 Mongooses exhibit a wide range of variation in such characters 

 as the numbers of the digits, the hairiness of the tarso-metatarsal 

 area, and the presence and extent of the interdigital webs. These 

 characters are either invariable or subject to much less variation 

 in other groups of corresponding rank amongst the -<951uroidea. 

 There is only one group of the suborder, however, which possesses 

 feet structurally lecalling those of the Mongooses, nameh% the 

 Galidictinse, the feet of which I have recently described and 

 figured *. Between the feet of Galidiciis and one of the penta- 

 dactyle, semiplantigrade Mongooses, like Miingos, thei'e appear 

 to be only two difierences which call for notice. In Galidictis 

 (and in Gcdidia) the pollex and the hallux are set lower on the 

 foot and project therefrom on a level with the internal lateral 

 lobe of the plantar pad, and the pollical and hallucal lobes of this 

 pad are better developed and in contact with the internal lateral 

 lobe. Hence the plantar pad is quadrilobate, Avhereas in Mungos 

 and all other genera of Mongooses the plantar pad is trilobate. 

 It may also be added that the metatarsal and carpal pads in 

 Gcdidia aiid Galidictis are better developed than in the Mongooses 

 and are double. 



These differences are interesting because they show that the 

 feet of the Galadictines are of a more primitive type and, on the 

 whole, more Viverrine than are those of the Mongooses. Never- 

 theless it cannot be claimed either that the feet of Mungos difler 

 more from those of Galidictis than they difier from the feet of 

 Bdeogale, Atilax, or Suricata, or that the feet of Galidictis differ 

 more from those of Mungos than they difler from the feet of the 

 Paradoxurine genera or of Eiiijleres. 



The Glandidar Anal Sac. 



The presence of a glandular anal sac in Mongooses lias long 

 been known ; but its invariable occurrence within the group has 

 been disputed. I have found it without exception in all the 

 specimens I have examined, even in those belonging to species 

 in which its existence has been denied. Cuvier, for example, 

 said that the Marsh-Mongoose, which he named Atilax vansb-ef, 

 is without it. It happens, on the contrary, to be ratlier ex- 

 ceptionally well developed in that form (text-fig. 9, B, C ). It is 

 also present, though small, in JIungos auro2)unctatus, despite 

 Mivart's statement + that in a living example he examined " the 

 ■anus opened most distinctly on the surface of the body, and not 

 into a saccular depression." Since Mivart was probably the 

 authority for Blanford's declaration § that "this character is 



* Ami. Mag. Nat. Hist. (8) xvi. pp. 351-356, pis. xiv., xv., 1915. 



t St. Hilaire & Cuvier, Hist. Nat. Mamiii. ii. pt. 54, pi. 198, 1826. 



X Proc. Zool. Soc. 1882, p. 178. 



§ ' Fauna of Brit. India ' : Mammalia, p. 119, 1888. It is a pity Blanford did not 

 particularise the species, .and saj' whether his information was hased upon his own 

 observations or not. It may here be recalled that Murie and others entirelj' failed 

 to find the large anal sac in a living Spotted Hysena. 



