194 THE CLASSIFICATION OF THE OPHIDIA. 



features. It remains to be seen whether the family I termed the Nothopidse, but 

 which Boulenger unites with the Acrochordida?, agrees with it in pulmonary characters. 

 The remarkable tracheal lung- or gland distinguishes the Epanodonta from the Cato- 

 donta, emphasizing the differences observed in the osteology of the skull. The huge 

 diverticulum of Heterodon serves to distinguish the genus from its allies. The extra- 

 ordinary transverse dilatation of the trachea in Thrasops establishes the genus as 

 distinct. 



The value of the rudimental right lung as a character of the Colubroidea is 

 increased by my investigations. In only two genera have I found it present or absent, 

 viz., Halsophis and Pityophis. I am not sure but that I may yet find it in the P. 

 melanoleitcu->, where I have failed hitherto, but I am sure that it is present in some 

 species of Halsophis and wanting in others. A natural group of American Colubrinse 

 appears to be characterized by its absence, viz., Rhinochilus, Cemophora and Ophibo- 

 lus, all genera with an entire anal shield. The development of cartilages in the bron- 

 chial foramen or tube of the rudimental lung, is not a constant character. I found it 

 in one Helerodon platyrMnus and not in another ; it is present in Conoi)Ms jjulcher, but 

 absent in C. sumiclirastii. 



The numerous characters presented by the hemipenis have various values. Sev- 

 eral very distinct types are distinguishable, but they are continuous at some point 

 through intermediate forms. This is, however, the history of all characters which 

 distinguish organic beings, especially of those which have been relied on as characters 

 of the minor divisions and genera of the Ophidia. The characters which I have dis- 

 covered in the hemipenis have added greatly to our resources in the attempt to learn 

 the relationships and hence origin of the members of the Ophidia. 



In a broad way we may distinguish as leading types the following : The smooth, 

 the plicate or flounced, the calyculate or ruched, and the disc-bearing. Any of these 

 may have the sulcus spermaticus simple or bifurcate, and they may have the middle 

 part of the organ spinous or not. The spines may extend to the apex so as to oblit- 

 erate the pattern, and the total organ may be bifurcate or not. As regards the indi- 

 cations of affinity presented by these types, it may be said that the nearer we 

 approach the Lacertilia the less spinous is the organ, and the farther away the form 

 the more certainly will the ruched structure prevail. The tendency to bifurcation is 

 present in most groups, but it is universal in but one suborder, the Solenoglypha, or 

 specialized venomous snakes. 



In the Oriental region we have the smoothest type of Colubroidea, which includes 

 the genera really allied to Calamaria, many of which have had hitherto widely differ- 

 ent positions in the systems. Owing to the scarcity of specimens of this type in 



