352 THE HON. r. A. METHUEN ON A NEW SNAKE, 



expanded into discs of moderate size, not quite as large as those 

 on third and fourth fingers ; a sliced pear-shaped inner meta- 

 tarsal tubercle, barely ^ length of first toe ; a small outer 

 metatarsal tubei'cle. 



The tibio-tarsal joint of the adpressed leg reaches a point 

 between the eye and the nostril. Heels strongly overlapping. 

 Length of tibia ^ the distance from snout to vent, about 3| times 

 as long as broad. 



Upper surfaces of head, body, and limbs very granular with 

 numerous warts and elongated tubercles, which on the head and 

 back form an irregular pattern. Belly, sides, and inner parts of 

 thighs granular. , 



Habit like the M. granulatus section of the genus Mantl- 

 dactylus (1). 



Colour : Above dark bluish-bi'own with irregular lighter mark- 

 ings more in evidence on the head than elsewhere ; upper lip, 

 loreal region, throat, and chest mottled with dirty white and dark 

 brown ; an irregular median dirty white line on throat and chest. 

 Upper surfaces of limbs same colour as back, the lighter brown 

 colour of the inner parts being carried across the upper surfaces 

 as thin irregular transverse bars. 



Length from snout to vent 27"50 mm. ; length of outstretched 

 hind limb from vent to tip of fourth toe 44 mm. 



Size of eggs in oviduct 275 mm. in diameter. 



Type a female, T. M. Oat. Kept. No. 1013, and cotype No. 1012, 

 in the Transvaal Museum ; origin, Folohy, East Madagascar, 1911 . 



In 1913 Mr. John Hewitt and I placed these two speci- 

 mens — for lack of comparative material — pro tern, in the genus 

 Mantidactylus, the genus to which this new form is most neai-ly 

 related (2). 



Trachymantis, gen. nov. 



Hemimantis : H. horrida Bttgr. Zool. Anz. 1880, p. 282 ; 

 Abhand. d. Senck. GeselL, B. xii.^p. 492, Taf. iii. fig. 14, 1881. 

 Arth'roleptis : A. horridas Bonleng. Cat. Bat. p. 118. 

 Microphryne Methn. & Hwtt. (2) p. 55, preoccupied. 



The only character I have to add to our original diagnosis 

 concerns the discs of the digits, as in Mantidactyhis and Gephy- 

 romantis the lower surface of the digits has the characteristic 

 I'ing-shaped groove. This character in Mantidactylus was pointed 

 out by Mr. Boulenger (1) when he made the genus Aglypto- 

 dactylus for Mantidactylus -niadaf/ascariensis. I must also correct 

 what was our impression at the time, that Trachymantis was 

 related to Rhacop)horus ; Gep>hijromantis, which is \evj closely 

 related to Mantidactylus and may even be a coixiparatively recent 

 offshoot from a M. gramdatiis-like form, leads naturally to 

 Trachymantis ; the relationship can be shown thus : — 



Mantidactylus : vomerine teeth ; outer metatarsals separated, 

 Gep)hyromantis : vomerine teeth ; outer meta.tarsals united. 

 Trachymantis : no vomeiine teeth ; outer metatarsals united. 



