Johnson — Axial Bifurcation in SnaJces. 533 



v.hicli is the' onlv case of the kind on record and which lies 

 without Fisher's system of classification. 



Degree of Bifurcation. The data as regards degree of bi- 

 furcation aside from the tliirteen cases here described and that 

 of Borgert have little quantitative value, for frequently the 

 descriptions are no more specific than merely '^two-headed." 

 Fortunately, however, the fact is sometimes noted whether 

 there are two heads on separate necks, or two skulls in organic 

 fusion. Classifying on this basis we have : 

 8 dipropi : 



7 near the eye. 



1 in the region of the nose. 

 23 dicephali: 



15 ''short neck." 



8 "long neck." 



In Tables I and II, and Plate XXXI, the degrees of bifurca- 

 tion of the thirteen snakes here described and the one of Borgert 

 (Xo. 14) are represented. Only by such precise criteria can the 

 question of degree of bifurcation be properly studied. Three 

 criteria were employed as a discontinuity might be shown by 

 one criterion which Avould not be shown bv the others. These 

 were the numl^er of vertebrae, the lengths of the divisions in 

 terms of proportion of the total length and in terms of vertebral 

 units. The lengths of the divisions were measured to the 

 cephalic end of the skeletal division, not the external division. 

 The verteljral unit v/as chosen, i. e., the length of one vertebra 

 where the vertebrae are largest, because it was found that occa- 

 sionally in the divisions there was a very large number of ab- 

 noiTnally small vertebra?. In each case, both classification by 

 the longer division and by the average, division was used. 



Even the few cases here described are so discontinuous in 

 their distribution as to degree that I Ijelieve it is safe to say 

 that axial bifurcation is more likelv to occur in some dearrees 



f o 



than in others, and that the most frequent degree is from about 

 6 to 13^ of the length from the cephalic end. 



Angle of Bifurcation. In the case of dicephali, the general 

 rule may be made that when the vertebral column is double for 

 a very short distance, the angle of divergence of the proximal 



