338 JOURNAL, BOMBAY NATURAL HlSTOR'^ SOCIETY, Vol. XlV. 



better means of differentiation in such cases, and that could simpler methods 

 be discovered from external characters alone, these would be welcome and 

 employed in preference ; but one has only to know a little of natural history 

 matters to realise that to discover such means is much more easily said than 

 done. It is with this object in view that I have been persuaded in this 

 paper to endeavour to simplify the means at our disposal of differentiating 

 between snakes, hoping that the results of my observations may be of use to 

 those who, by reason of their access to large and fully representative collec- 

 tions, have it in their power to compile works — a possibility forbidden to one 

 in my position who only sees at most a limited number of species and genera. 



Many of the points I have made reference to may be found of doubtful 

 utility in some genera, but the same may be said of all scales, and this should 

 net detract from their value in others. Even such important scales in 

 diagnosis, as labials for instance, have not an equal value in all genera. In 

 Bungarus and Naia I have never yet- seen one departure from the nor- 

 mal, whereas in CJirysopelea and many Tropidonoius, &c., these are not 

 infrequent, and in many Hydrophince. these scales are particularly unreliable 

 owing to their inconstancy. Similar instances may be cited with reference to 

 almost every scale. 



I have adopted the nomenclature in usage in Boulenger's work, " Fauna of 

 British India" — "Reptilia and Batrachia," 



The Dorsals. 

 These scales are perhaps the most useful of all guides in enabling us to 

 differentiate between species and genera, but authors have not availed them- 

 selves of their full value. It is usual for them to record the number of the 

 rows in the middle of the body and at this point only, but observation has 

 revealed to me that information as valuable is to be obtained by counting 

 the rows in other situations as well. The rows which can be counted from 

 the ventrals on one side over the back to the ventrals on the opposite side 

 in far the majority of snakes total in the aggregate an odd number. In the 

 neck these are more numerous than in the body, and these rapidly decrease, 

 so that very shortly after the neck they dwindle to a definite and constant 

 number in like species, though it may be very different in different species. 

 This number once established is retained for a variable distance in the length 

 of the snake. As far as my observations serve me, the number is preserved 

 to a point at least well behind the middle of the body (not including the tail^ 

 which begins at the vent), after which the arrangement differs according to 

 the species. In many cases the number peculiar to the species is preserved 

 in the whole length of the body. On the other hand, in many other snakes 

 the number peculiar to the species once established remains so to a somewhat 

 variable point behind the middle of the body, after which, by the absorption 

 of a row on each side, the number is diminished by two, which number 

 remains constant to the vent, or in some cases a second or third absorption of 

 rows may occur before the vent. These steps occur at intervals which are' 



