COLLECTION OF SNAKES IN THE NILGIRI HILLS. 557 



They agree very well with the description given in Boulenger''s Catalogue 

 (Vol. I, p. 26) except that my specimens are much more slender, the diameter 

 of the body being respectively about 1/75, 1/64, and 1/82 the total length. 



My specimens measured 12^, lOj and 9| inches. 



I give two drawings of the head shields. 



Family VROPELTID^. 

 Beddome's Shieldtail. — RhinopMs sanguineus (Beddome.) 



All my specimens, 40 in number, came from the Wynaad where it is 

 evidently a common species. 



Food. — Many contained fragments of earth worms in the stomach, and 

 the intestines were nearly always loaded with semi-liquid mud from their 

 victims' alimentary systems. The fact that the worms were found in 

 fragments suggests that when seized they break themselves free by their 

 struggles, only to be recaptured, and repeat the process. 



Sexes. — Of 35 sexed 16 were $ and 19 d" ■ The following sexual differ- 

 ences were noted. In females the body is rather longer and the tail shorter 

 than in the male. This is seen by the ventrals in the former ranging from 

 214 to 218 against 200 to 213 in the male, and the subcaudals numbering 5 

 to 7 in the female against 9 to 11 in the male. In the male also the last 

 ventrals, the last scales in the lowest 3 rows of costals, the anal, and the 

 subcaudal shields are pluricarinate. The keels are rather indistinct, and 

 only seen on a subterminal zone on these shields (see figure). There is no 

 trace of these keels in the female. 



Breeding. — Although I got no gravid $ , many juvenile specimens with 

 open navels prove that the young embark on life late in July, August, and 

 September. 



Growth. — No less than eleven were young of the year, and varied in 

 length from 4f to 5| inches. No specimen between 5| and 10 inches was 

 obtained so that it would appear that the young double their length in the 

 first year of life. Nearly all the specimens were from 11 to 13 inches 

 long. One $ measured 16 and one c? 14 inches. The rule m Colubrines 

 is for the young to be about one-foui'th the average adult measurement. 



Colour. — No young specimens were brightly marked with coral-red ven- 

 trally. Only a faint tinge of pink was seen until adult life in this region. 



Lepidosis. — In most specimens several of the subcaudal shields were entire. 



The skin strips off this snake as easily as from others, except on the end 

 of the tail. Here the modified skin on the terminal scute is so intimately 

 adherent to the terminal vertebrae, that it is only removed with consider- 

 able difficulty. 



Dentition.— The maxillary teeth number 5. There are no teeth in the 

 palatine, and pterygoid bones. The mandibular set number 5. 



The Argus Roughtail.— 6Y7y6Mra ocellata (Beddome;. 



This is an even commoner species than the last in the Wynaad (including 

 Liddelsdale), my aggregate being 101. No specimen reached me from any 

 other locality. 



Food. — This consists entirely of earth worms, and the remarks made 

 under the diet of the last species apply equally well to this snake. 



Disposition. — I had a few live examples sent me, and I found them all 

 very inoffensive creatures. They are rather restless in one's grasp, and 

 push their noses through and through the clefts of the fingers. I put them 

 on to loose earth, and found sometimes they would burrow, and at other 

 times they remained inert, making no attempt to glide away. They move 

 very slowly. Those that burrowed used the snout only for this purpose, 

 and 1 am still perplexed as to the use of the curious tail. This is so often 



