1908.] OF SNAKES OF THE GENUS CORALLUS. 153' 



may or may not be a moderately long and thin posterior prolonga- 

 tion of one of the lobes ; the intrapulmonary bronchus also is.- 

 longer and shorter in some specimens than in others ; the spleen 

 is by no means entirely constant in its form in every individual. 

 The position of the kidneys varies slightly, as does the amount of 

 overlap of one by the other. The table showing the relative 

 position of the different organs of the body indicates that there 

 is, as might be expected, a correspondence between the leng-ths 

 of the snakes and the distances separating the viscera. The 

 table, however, does not show satisfactorily where the growth of 

 the individual takes place, except that it would perhaps be 

 inferred that it is not in the region following the liver ; for in 

 two individuals measuring respectively 63 and 48| inches the 

 distance from the end of the liver to the gall-bladder is precisely 

 the same. I observed this carefully in the two individuals F 

 and G. There is a difference of ten or eleven inches in total length 

 between the snakes, and yet the distances between the end of the 

 liver, the spleen, and the gall-bladder were exactly the same, as 

 was conclusively shown by placing the snakes side by side, when 

 the organs mentioned were seen to correspond exactly in position. 

 This seems to show that the growth in length takes place both in 

 front of and behind this particular region of the body. I should 

 imagine that this region corresponds to the umbilical region of 

 the new-born snake. In conclusion, it may be pointed out that 

 the comparative fixity in the relative position of organs in so 

 many specimens and of different lengths confirms the use of thia 

 character as of systematic value in snakes. 



§ Conclusions. 



It will be clear that the characteristics of these two species of 

 Corallus, even if they be ultimately referred to two distinct, 

 genera, breaks down my previous attempts to define the sub- 

 families Boiiice and Pythonince by anatomical characters, and 

 renders those charactei'S only generally applicable. Hitherto 

 these two subfamilies have been distinguished by the existence in 

 the Pythonince of a supraorbital bone wedged in between the; 

 prefrontal, frontal, and postf rental bones ; this bone is wanting 

 in the Boince. I presume that Corallus has been examined in 

 this respect, and has been found to possess no supraorbital bone. 

 Otherwise the Boince (and Corctllus cookii) agree to differ from 

 the Pythonince, (including Corctllus maclagascariensis !) in the 

 mode of distribution of the inteixostal arteries, and in the 

 persistence of a considerable umbilical vein, I am disposed to 

 think that in view of facts accumulated since the division of the 

 Boidse as indicated above, it is not so desirable for the present to 

 insist upon any such subdivision, which is, after all, by no means 

 in accord with geographical range. On the other hand, the 

 present communication enables me to substantiate still further 

 the characteristics of the family Boidfe, which differs from all other 



