387 



The present series, kindly loaned me by Dr. Barbour, and all 

 collected by myself, includes: M. C. Z. no. 7949, 7951-2 from 

 Monteverde, M. C. Z. no. 8040 from Zent, M. C. Z. no. 9785-7 

 from Suretka, M. C. Z. no. 9879, 9882-4 from Almirante, and 

 M. C. Z. no. 9971-5 from La Loma, in all 16 specimens, and the 

 four distributed specimens from the original lot have been sent 

 out in exchange, so that I have seen at least 22 of this quite 

 distinct species. 



I took the type with me to various European Museums so 

 that it has been compared directly with the types of nubilus, 

 fitzingeri, longirostris, conspicillatus, raniformis, etc., and it 

 agrees with none of them. 



Lithodytes gaigei, sp. no v. 



Type—M. C. Z. no. 10011. 



Type locality. — yort Randolph, Panama Canal Zone. 



Range. — Known only from Fort Randolph and Barro Colorado Island, Pan- 

 ama Canal Zone, and from Talamanca Valley, Costa Rica. 



Diagnosis. — Close to Lithodytes lineatus of South America, but differing in 

 having head wider than body; no flash markings on thigh or in groin; vomerine 

 teeth obhque rather than transverse. 



Description. — M. C. Z. no. 10011: head a rather pointed oval; snout flat 

 above, pointed; canthus rostralis angular; lores sloping steeply; upper eyelid 

 less than interorbital diameter; eye equal to its distance from nostril; tym- 

 panum }/2 eye; heel to between snout and eye; head wider than body; disks 

 of fingers and toes equal, about half again diameter of digit, much smaller 

 than tympanum; fingers and toes free; a large inner and small outer meta- 

 tarsal tubercle; a tarsal fold; finger I about equal to finger II; toe III equal to 

 toe IV; smooth below, uniformly and finely rugose above; vomerine teeth in 

 two oblique groups, equally in and back from choanae, separated by diameter 

 of choanae and distant from choanae by same distance; dark brown, lighter 

 below; a Ught line from eye to sacrum, snout to vent 38 mm. 



Variation. — A specimen from Barro Colorado Island, Univ. 

 Michigan field no. 77, is much hghter below, and the dorso- 

 lateral lines are hard to make out. Also the tarsal fold is in- 

 visible (the animal is soft); the first finger on one side is longer 

 than the second and is shorter on the other; the third toe is 

 longer than the fifth; snout to vent 39 mm. Two small specimens 

 from Talamanca, M. C. Z. no. 9901, 9904, are 21 and 18 mm. 

 in length. They are black above, slightly lighter below, and 

 the dorso-lateral light lines begin on the snout where they meet. 

 The vomerine teeth cannot be made out. In one the first and 



