13 



the point equal to about one-tenth of the length of the 

 fang ; a groove connects the orifices, or rather did connect 

 them during the growth of the fang, at which time the 

 canal, originally open in its entire length, became closed. 

 The canal only occupies the front of the fang ; the hinder 

 part is a bony column giving considerable strength to the 

 structure. 



In the Viperine snakes a transition takes place, gradu- 

 ally culminating in the most perfect form of poison- 

 apparatus, viz., a long fang usually lying supine along the 

 jaw, but capable of full erection The genus Trime- 

 surus is not nearly so complete as this, the fang is long, 

 but the erection imperfect ; the maxillary consists of an 

 open shell communicating with the exterior of the cheek and 

 forming the pit characteristic of the crotaline snakes. But 

 it is in Daboia that we see the perfection of mechanism ; 

 on removal of the skin covering the cheek, we come at once 

 across the strong jugal ligament lying below the eye ; it 

 binds the upper part of the maxillary to the prefrontal and 

 postfrontal bones, thus affording a fulcrum for the action 

 of the erectile apparatus. The maxilla is found to be con- 

 siderably modified in form ; it is no longer placed below the 

 orbit, this position is occupied by the elongated ecto- 

 pteiygoid, whilst the maxillary, only one-fifth of an inch long 

 (in a large Daboia) but double that in height, is placed at 

 the end of this bone like a hammer-head at the end of its 

 handle. 



Imagine a hammer-head with the claw downwards (repre- 

 senting the fang,) and hinged at its junction with the handle 

 (the ectopterygoid bone). Now if the top of the hammer- 

 head be fixed by a ligament to a fulcrum, protrusion or 

 retraction of the handle will cause the claw to be either 

 erected or depressed. 



The muscles by which the ectopterygoid bone is thus 



