DORSAL SCALE-ROWS IN BRITISH SNAKES. 363 



the prey whole. What would be the use of the highly elastic 

 ligamentous attachment of the mandible, and the loose attach- 

 ment of the bones of the skull generally, if the skin over the 

 back of the head and neck were not also capable of the highest 

 distension ? The scale-rows appear first and last upon this small 

 area, but the development of fresh rows almost invariably halts 

 on the neck, and recommences over the region of the stomach, 

 afterwards continuing forwards and backwards until the scale- 

 series is complete from the head to the level of the pyloric 

 sphincter — the second halting-point. It is probable that this 

 procedure in the production and reduction of scale-i'ows will be 

 found to hold e'ood for all snakes. 



Sexual dirno)-pli is)n. 



In accordance witli these principles, it follows that the female 

 snake should have, or tend to have, a higher number of scale- 

 rows than the male, on account of her greater girth. Dr. Ruthven 

 has already made this supposition*, but finds that it is not 

 realised in the Garter-snakes, except in the case of T. radiw, in 

 which the greater number of males have the formula 19-21-19-17, 

 whilst the greater number of females have that of 21-19-17. 



In T. natrix both sexes have the formula 19-17, and I can 

 detect no sexual variation in the amount of abridgment of 

 series IV. in the 25 British specimens which I have studied. 



All the British specimens of C. austriaca in the British 

 Museum have the formula 19-17-15. It may be that there is a 

 sexual variation in the position of tlie posterior reduction, as in 

 8 out of 9 females this occurred from "85 to "96 in the length 

 of the body, whilst in 3 out of 4 males it occurred '75 to -85 in 

 the length of the body ; one cannot base conclusions, however, 

 on such small material as this, or the preceding. 



In V. herus there is undoubtedly a tendency for the females to 

 possess a higher formula than the males. The number of rows 

 on the middle third of the body varies from 19 to 23. In 

 170 British specimens the 4 having 28 rows were all females. 

 Mr. Boulenger, who ha,s statistics relating to 460 European speci- 

 mens, finds that 3*74 per cent, males and 2-81 per cent, females 

 have less than 21 rows, and 2-79 per cent, males and 6-32 per 

 cent, females have more than 21 rows. This tendency is still 

 more marked in the Green Viper, Atheris squamiyer f Hallow., in 

 which the male has from 19 to 21, and the female 21 to 23 rows. 



At present there is only one species known, in which this 

 difierentiatiou between the sexes is complete and constant. Mr. 

 Boulenger J has recently discovered that in the Cameroon Bothro- 



* Op. oil. 



t Boulenger, " Batraciens et Reptiles recueillis par le Dr. Christy au Congo 

 Beige 1912-1914." Revue Zool. Afi-ic. vol. vii. fase. 1, 1919. 



X " Un cas de dimorpliism sexual chez un Serpent Africain {BothfoJycus citer 

 Giintli.)." Inst, de France, C. R. Ac. Sci. 1919, 



