9 



important it is often a fallacious guide so that all the other 

 points should be considered first. 



The colour resides chiefly in the scales but the skin is 

 often coloured too and this shows up between the scales, 

 forming what is known as the interstitial colour. The 

 margins of the scales may be coloured differently to the 

 scales themselves. In fact a scale may be of several different 

 colours just like a bird's feather may. All sorts of compli- 

 cated patterns are thus produced — stripes, dots, ring- spots, 

 ocelli or " eyes," transverse and longitudinal lines and 

 bars, etc. Hence it is very difficult to accurately describe 

 in words or to paint a snake's colouring. In fact coloured 

 pictures of snakes are, I find, of very little use in identifying 

 an unknown snake. Still there is generally a main ground 

 colour and some fairly definite sort of pattern so that, 

 although it may be difficult to imagine a snake from its 

 written description, still with the specimens before one it can 

 generally be recognised from its description. All the same, 

 colour is a variable characteristic and should never be relied 

 on alone to identify a snake. 



CHAPTER III. 



VARIATIONS FROM THE TYPICAL SNAKE AS JUST DESCRIBED. 



As we saw in the last chapter the typical snake is 

 covered with comparatively small scales on its body, shields 

 on its head and ventrals on its belly. 



But there are certain variations from this typical arrange- 

 ment which we have to study. 



Turn a typical snake over on to its back and we have 

 seen that the ventrals are transverse plates extending right 

 across the belly (Figs. 4 and 5). But in some snakes the 

 ventrals are much narrower than this. So that as the snake 

 lies on its back we can see the whole of the last row and 

 generally several rows of the ordinary scales on each side of 



