1920.] Dr. Malcolm Smith : Sea Snakes. 3 



Some good collections made well away from the coast 

 would be of value in helping to elucidate this interesting 

 problem. 



This eccentricity of distribution, as well as minor varia- 

 tions in scalation and colour which I have found confined to 

 certain localities only, confirms the view that I have held 

 for some time, namely, that although the range of a species 

 may be very great, that of the individuals comprising it is 

 inclined to be extremely local. 



It follows from this that we may expect to find among 

 sea snakes a considerable number of geographical forms. 

 No attempt has yet been made to define races for any of the 

 species, for the reason that few herpetologists have had 

 sufficient material to work upon. Yet it is clear, when a 

 good series can be obtained from one locality, and compared 

 with a series from another, sufficiently remote, that 

 differences can be found. In many cases the difference 

 is not great, and is confined merely to slight variation in 

 the number of scales round the neck and body. A few 

 shew more distinct changes, not only in the number of 

 scale rows but in the relationship of the head shields to 

 one ani^her. Owing to the wide variation which the 

 indivicl'uals of a species may exhibit in any one locality, 

 large series are naturally required before the range of 

 variation of any particular race can be defined. It is 

 imperative also, if satisfactory results are to be obtained, 

 that the same methods of examination should be used. 

 For, according to the way in which the scale rows and 

 ventral shields are counted, so will the results differ. 



In counting the body scales I have endeavoured to find 

 the minimum and the maximum number of rows. This 

 necessitates several counts at each spot. The minimum 

 is upon the neck, usually from two to three, but sometimes 

 as much as from four to five, heads-lengths behind the 

 head. The maximum is at mid-body or posterior to it. 

 In certain forms, such as Enliydris, Thalassophis, Hydras, 

 the maximum is usually attained by mid-body, but in most 

 of them, and particularly in (he small headed forms, it 

 is not reached until well pasi mid-body. The ventrals are 

 counted from the first biluberculated shield that can be 

 discovered upon the neck, and all missing ones are allowed 

 for as if they were present. 



The range in the number of ventral shields varies 

 considerably. It is greatest in that group of species which 

 are at present defined under " head very small, body very 

 long and slender anteriorly." It will be seen however, upon 

 inspection of the tables given, that while the range of the 

 majority in a series is comparatively limited, a few indivi- 

 duals at cither end string it out and add greatly to its 

 number. It is disturbing, after having examined a large 

 series, and obtained what appears to be the range of varia- 

 tion of these shields, to find another example which upsets 

 one's figures to the extent of twenty, thirty, or even more. 

 Some specimens in my collection differ in this respect so 



