VARIATION OF THE PIT-VIPER» 167 



that of L. atrox is of a " tr^s beau jaune clair marquete sur les 

 bords de noir ou de gris fonce"*. 



In examining a large number of these snakes I have found the 

 ground-colour to be equally variable in both ; the ventrals may 

 be yellow, yellow checkered with black, or of that greenish colour 

 produced by a powderiiig of black over yellow, in either L. atrox 

 or L. lanceolatus. It was ther-ef ore not quite clear to me why 

 Dr. Brazil considers the general aspect and coloration such striking 

 points, or how the latter should produce a velvety appearance in 

 L. atrox. Mr. Boulenger has looked into this matter with me, 

 using specimens of L. atrox and L. lanceolatus named and sent to 

 him by Dr. Brazil himself. There certainly is a difference in the 

 appearance of these specimens, but this is not due in any way 

 to the markings, which are strikingly similar. The specimen 

 marked " at7'ox" has, however, the very prominently keeled scales 

 mentioned before, and Mr. Boulenger has pointed out to me that 

 the " velvety " appearance of this snake is entirely due to the 

 consequently greater breaking up of light upon its dorsal surface. 

 Dr. Brazil's L. atrox is therefore the same as Mr. Boulenger's. 



L.jararacussu. — Dr. Brazil maintains the specific rank assigned 

 to the Jararacussu by Lacerda : first, on account of its striking 

 black and yellow markings ; secondly, because of the comparatively 

 low number of its ventrals ; and, thirdly, because of its much 

 more triangular head. With regard to the first point, there are 

 specimens in the British Museum collection which can be graded 

 to form a transition series from the typical L. atrox to the 

 Jararacussu, a specimen from W. Ecuador being exactly inter- 

 mediate between the two types of markings. With regard to the 

 second point, the ventrals of the Jararacussu are as a rule fewer 

 than those of L. atrox. Dr. Brazil mentions 170-176, but two 

 specimens received from him have 180 and 184. 180-240 ventrals 

 is, however, the range given for L. lanceolatus in Mr. Boulenger's 

 Catalogue of Snakes, and, as I have failed to detect any structural 

 differences in the shape of the head or in the scaling, I think 

 that the Jararacussu cannot be regarded as more than a colour 

 variety of L. atrox L. 



II. FORBI AND LePIDOSIS. 



As I think I have shown in the annexed table that there is no 

 correspondence between variations in markings and the number of 

 ventrals and caudals, I will describe the differences in form and 

 lepidosis which have led to the distinction of species. 



The most important is that of the two types of scaling found 

 in L. atrox. It has been generally accepted that the high short 



* The coloured plates which accompany Dr. V. Brazil's descriptions do not convey 

 these supposed difierences, and it is well to point out that but little care has been 

 bestowed on the rendering of the markings. I think a L. lanceolatus with only 

 about 10 markings on the body, as figured, to be an impossibility. 



