34! 



is the usual colour of tree-snakes ; it is very delicate, and 

 rapidly passes through changes from tender green to bronze 

 and blue. In a few cases the colours are dead, but generally 

 they are shining, even iridescent in certain lights, and afford 

 a beautiful play of colours. 



The patterns in which these colours are arranged are 

 often very difficult to describe, and it is by no means easy 

 to imagine the actual pattern and colours of a snake from 

 a verbal or written description However accurate. The 

 entrance of an interstitial pattern from the skin below, the 

 secondary patterns produced by dark tinged margins to the 

 scales, and the play of colours in different lights, sorely tax 

 the word-painter's power of description. 



Th^ patterns are formed by stripes and by series of dots, of 

 ring-spots, ofocelli or of other shaped marks in a longitudinal 

 or in a transverse direction, or in both. The longitudinal lines 

 may cross the transverse lines or vice versd, and the points 

 of crossing may be marked by a different colour. Cross- 

 bars are often ocellate, that is, including eyes in their course, 

 and a fasciolated pattern is common ; it consists of cross- 

 bars of variegated colour produced by darker or lighter 

 tips to certain series of scales. It frequently happens that 

 cross-patterns are unsymmetrical, the bars of either side 

 not meeting exactly at the median line. 



Cross-markings rarely extend all round the body, except 

 in a few snakes encircled by rings ; generally the under-parts 

 are of a different pattern, plain, mottled, banded, or spotted. 

 The throat is also of a different colour, lighter, often yellow. 

 The head is sometimes marked with fillets, and streaks 

 frequently pass downwards or obliquely backwards from the 

 eye to the throat. Collars are very common, either =» shaped 

 (point forward) or ■<; shaped (point backward). 



The exact period at which snakes cast their skins is very 

 variable, but about two months appears to be the averat^e 



