8 
INTRODUCTOEY LETTER. 
should have, in all its parts, a mathematical exactness, so 
that one could study the relative proportions of the organs 
it is for the same reason necessary to avoid fore-shorten- 
ings as much as possible, and not to confuse the figure by 
projected shadows. The choice of objects demands equal 
circumspection : the agony of a violent death often leaves 
traces in the convulsed features of the face ; some parts 
sustain accidental injuries in putting up the specimens ; 
in others, the organs change their relative positions. This 
is especially the case with Ophidians, of which the bones 
of the head are susceptible of very considerable movements. 
In tracing my figures, I have aimed at avoiding all these 
obstacles : I have always made use of individuals in a 
perfect state of preservation, and I have never copied, ex- 
cept from nature. I have followed the same rules in my 
descriptions ; or, when I have not been able to do so, I have 
expressly indicated it. 
The comparative examination of the writings of my 
predecessors, presented far more difiicult obstacles to be 
surmounted. The principal cause, to which I have fre- 
quently alluded, and which appears to have engendered 
the numerous errors that disfigure our systems, is the 
multiplication of species and of genera instituted and in- 
troduced into systems upon isolated characters. The 
instances that would justify these remarks are innumer- 
able. Let any one consult my articles Eryx, Naja por- 
phyritica. Boa Cenchria, Python Peronii and P. bivittatus, 
Acrochordus, Tropidonotus bipunctatus, and T. fasciatus, 
Crotalus horridus, Vipera Berus, and several others, and 
he will be convinced that the same species often bears a 
dozen of names ; that it often has been divided into several 
different genera; that it has even formed the types of 
different families ; that it figures sometimes among the 
venomous serpents, sometimes among those that are in- 
nocuous ! Let any one examine the heterogeneous elements 
out of which some authors have composed their genera 
Boa, Hurria, Scytale, Elaps, Trimeresurus, and Vipera ! 
In following out such views, it was necessary, in adhering 
to this principle of classification, to separate the Pythons 
from the Boas, and to place them in another family ; it 
