450 PROCEEDINGS OF THE NATIONAL MUSEUM. vol. xxxvi. 



tAvo genera are very distinct. The premolars both above and below 

 in Hylomys are f, in Podogymnural. The three anterior premolars 

 in Ilylomi/s are small, practically functionless teeth, while in Podo- 

 gymnura the penultimate premolar is well developed and trenchant, 

 standing nearly as high in the tooth row as the last premolar (Plate 

 36) . The premolars in Gymn ura are f , as they are in nylo))tys. The 

 two anterior of them are small and practically functionless, like the 

 three anterior premolars of Hylomys; the tAvo posterior are large 

 and trenchant, like the two j^osterior premolars of Podogymnura. 

 (Compare Plates 35 and 3G.) Of the three genera GyinnuTa appears 

 to be tlie most generalized and primitive, and from it or its ancestors 

 the other two genera have been derived. In the shortening of the tail 

 and simplification in the structure of the premolars Hylomys has 

 depai'ted further from Gymnura than has Podogymnura. The last 

 upper molar of Podogymnura and Hylomys shows a more simple 

 structure than the corresponding tooth of Gymnura^ which has a 

 distinct posterior heel, lacking in the other two genera. The genus 

 Gymnura is said to present the most generalized structure known 

 among placental mammals." An account of its anatomy is given in 

 Dobson's Monogi'aph of the Insect ivora, 1882. 



The geographic distribution of the genus coincides with part of 

 the Malayan subregion of the Oriental region, namely, Tenasserim, 

 tlie Malay Peninsula, Sumatra, and Borneo, and a few immediately 

 adjacent islands. So far as known, it does not occur on others of the 

 larger or of more remote smaller islands of the Mala3^an Archipelago. 



Altliough the existence of tAVo species in the genus had l)een pointed 

 out by Giebel in 18(')3,'' yet they were usually considered as mere local 

 varieties, or color phases, until Jentink reaffirmed their specific dis- 

 tinction in 1881.*^ Jentink and other writers have described certain 

 peculiarities of shape in the skull and teeth by Avhich the two species 

 in the genus may be distinguished, but I have been unable to detect 

 any other peculiarities than size and color for distingnshing them. 

 The skulls and teeth, hoAvever. shoAv many individual A^ariations, but 

 none of them are constant for definite geographic areas. The char- 

 acters assigned by other writers may probably be explained as the 

 result of examining too small a number of specimens. 



The different members of the genus may be distinguished by the 

 following key : 



KEY TO THE SPECIES AND SUBSPECIES OF GYMNURA. 



n. Color uniformly Avbite or whitish, size larjiP, hindfoot 00-74 mm., bnsal 

 length of skull 70-84_l Gymmira alha, p. 453 



« Huxley, Proc. Zool. See. London, 1880, d- 657. 

 * Zeitschr. Ges. Naturw., XXII, p. 277. 

 '•Notes Leyden Mus., Ill, pp. 16G-168. 



