288 PROCEEDINGS OF THE NATIONAL MUSEUM. vol. 41. 



tubercular; canthus rostralis raised, granular; snout rounded, slightly 

 longer than the diameter of the orbit; crown and loreal region con- 

 cave; interorbital space much wdder than upper eyelid; tympanum cir- 

 cular, very distinct, nearly as large as the diameter of the eye; fingers 

 free; toes one-third webbed; disks large, about three-fourths diam- 

 eter of tympanum; subarticular tubercles moderate; inner metatarsal 

 tubercle prominent, no outer; a distinct tarsal fold; heels overlapping; 

 hind limb, being carried forward along the body, reaches anterior 

 angle of the eye; skin coarsely granular above, more distinctly so on 

 the abdomen and underside of thigh; throat and chest smooth. Color 

 (in alcohol) above dark bluish slate; on the sides and in the groin, 

 several large, vertical, dark spots, narrowly edged with whitish; hind 

 Kmbs with transverse bands of dark color similarly edged with wliitish; 

 underside paler than back. 



Dimensions. 



mm. 



Tip of snout to vent 60 



Tip of snout to posterior edge of tympanum 21 



Tip of snout to posterior edge of casque on the median line 18 



Width of head 22 



Fore leg from axilla 41 



Hind leg from vent 93 



Tibia 31 



Remarlcs. — The species here described is probably the same as the 

 young specimen from Carillo, Costa Rica, described and figured by 

 Guenther in the Biologia Cent rali- Americana^ as Nototrema ovife- 

 rum, and possibly also mth the Mexican specimen from Cordova in 

 British Museum, described by Boulenger^ under the same name. I 

 feel confident, however, that none of these belong to the true Gastrotheca 

 ovifera of Weinland,^ from Venezuela. In the latter the casque is in- 

 closed behind by a "tuberculous wall" of bone, the outhne of which 

 forms a wide median extension backward and two deep sinuses later- 

 ally; the vomerine teeth are between the choanae, and the fingers and 

 toes are much more extensively webbed. Moreover, the color pattern 

 is essentially different. 



The genus Gastrotheca was estabHshed by Fitzinger in 1843,* who 

 specifically designated Hyla marswpiata Dumeril and Bibron as the 

 type. Fifteen years afterwards Dr. Guenther ^ thought liimself "jus- 

 tified in not accepting his name," because from its supposed Greek 

 derivation "Fitzinger appears to have believed in the existence of a 

 pouch on the belly," and he accordingly changed it into Nototrema. 

 This, of course, is in contravention of all zoological codes of nomen- 

 clature, and Fitzinger's name has to be restored. 



1 Biol. Centr.-Amer., Kept. Batr., 1901, p. 288, pi. 74, fig. A. 



2 Cat. Batr. Sal. Brit. Mus., 1882. p. 418. 



3 Weinland, in Mueller's Arch. Anat. Physiol., 1854, p. 473, pi. 17. 

 <Syst. Rept.p. 30. 



6 Cat. Batr. Sal. Brit. Mus., 1858, p. 115. 



