420 ON THE GHOWTH AND STBUCTURE 
But it is carefully to be kept in mind, that this osseous 
tube has neither furrow nor fissure^ hut is at all times com- 
plete. It is probable that, in the mean time, the other 
tooth is detached, in order to be thrown off, by a process 
exactly similar to that by which the poison canal is formed ; 
but the process of decay and renovation, or the loss of one 
tooth and its replacement by another, does not seem to be 
in any way influenced by mechanical laws, or to be neces- 
sarily connected with each other as cause and effect, for I 
discovered and carefully preserved the head of a common 
viper, in which there are two fangs firmly implanted into 
one maxillary bone. It would seem that the replacement 
of the teeth takes place at fixed, though unknown, periods. 
The sections of the full grown tooth present certain differ- 
ences from those described in the rudimentary ones ; if we'^ 
commence, for example, near th© base, and cut the tooth 
across, in the situation of the opening at the base *, it will 
seem as if the poison canal opened on the convex surface ; 
but at every point anterior to this, or towards the point of 
the fang, the section will present a complete central hollow 
tube or canal, nearly circular, surrounded, excepting at one 
point, by a pulpy matter (itself inclosed by the external pa- 
rietes of the tooth), and therefore not continuous, being 
separated on the convex side of the tooth by an osseous 
vertical lamina, connecting the poison canal and external 
containing case of the tooth together. 
The rudimentary teeth are found to vary in number, 
which may probably depend on the age of the animal. In 
the common Viper of this country, and in the Puff- Adder 
of Africa, I have generally found five ; in other venemous 
serpents, four ; in the Rattle-Snake, five ; in some Cobras, 
* See the Figures in Plate X. 
