HEEPESTES. 171 



for these two processes to unite in adult life and to enclose tlie orbit posteriorly, 

 complete union of these processes generally taking place when the sutures of the 

 cranium are all amalgamated. I know of only one species of the Asiatic Herpestes 

 I have examined, in which these two processes probably do not unite, viz., H. semi- 

 torquakis, and in which the orbit is always imperfect. The infra-orbital foramen 

 is well defined. The palatal surface is triangxdar, with the palatines prolonged back- 

 wards beyond the alveolar border, as in the skulls of allied families, and with the ptery- 

 goids forming the lateral angles of the posterior nares, the commencement of which is 

 opposite to the articular sm'face of the squamous. The palatal border of the posterior 

 nares may be either simply arched, or the arched portion may contract to a narrow 

 notch, or the palatine border may be transverse. The tympanic bullse are prominent, 

 and project below and external to the opening of their osseous tube. The lower jaw 

 has a prominent, ascending ramus, and a somewhat upwardly reverted, backwardly 

 projecting, posterior angle. I have tabulated the measm-ements of the skulls of the 

 different species, and an inspection of the table and accompanying figures shows 

 how uniform the skuU characters are, and how closely some of the species are related. 



Dr. Gray included in the genus Herpestes, of which S. ichneumon may be taken 

 as the type, Herpestes jerdonii, S. pallidus, U. p)ersiGus, H. fuscus, H. javanicus, H. 

 semitorqnatus, H. exilis, H. malaccensis, and S. brachyurus. The foregoing is the 

 order piu'sued in the Catalogue of Carnivora, &c., in the British Museum, in which 

 the Madras Herpestes jerdonii, which is closely allied to H. pallidus, is placed along 

 with the African forms, although the genus Herpestes is divided by Dr. Gray into an 

 African and Asiatic sub-division. That this arrangement is thoroughly artificial is 

 evinced by the circumstance that the small Mungoose, H. persicus, which is apparently 

 only a local form of H. miropunctatus, is placed between H. pallidus and H. fuscus, 

 two large forms nearly allied to each other, but markedly distinct from H. persictis, 

 whose nearest ally, H. auropunctatus, is located in a separate genus Calogale, to 

 which Herpestes pallidus is also referred under Hodgson's specific name of nyula. 

 Again, H. exilis, Gervais, which is identical with H. rutilus. Gray, and H. Javanicus, 

 is classified along with H. seniitorquatiis and H. malaccensis, the former being a Mun- 

 goose nearly afiined to H. urva and H. vitticollis, while H. malaccensis is evidently 

 the same as H. piallidus. The Ceylon Mungoose, H. maccartliice, is ranked as a distinct 

 genus, Onychogale, on account of the length of its claws, which is undoubtedly a 

 result of the animal having been kept in confinement ; while the other Ceylon form, 

 H. smithii, which has a complete and not " rather incomplete" orbit as mentioned in 

 the foregoing Catalogue, is also placed in a separate genus Calictis, apparently also 

 founded on claw characters, the result of confinement. The large Madras Mungoose, 

 H. vitticollis, which has its nearest afiines in H. urva and H. semitorquatus, is 

 regarded as the type of a distinct genus Tceniogale, whilst H. urva has also its own 

 genus TIrva, and the Bornean, neck-banded Mungoose, which is closely allied 

 to these, is allocated to Herpestes. 



The genus Tceniogale is founded on altogether fallacious grounds, viz., the 

 supposed presence of 42 teeth. Ogilby, in 1835, described the skull, now in the 



