34 
introduce into the science numerous sorts founded on these 
materials, and he went so far as to consider as sufficient docu- 
ments rough paintings due to Chinese and Japanese draughtsmen. 
When he only used materials due to such men as Commerson, 
Forster, &c., he was pretty safe ; hut even then, it is well known 
that a naturalist, travelling in little known countries, is often so 
pressed for time as to necessarily neglect in his drawings many 
characters that will be found necessary when they will be sub- 
mitted to the scrutiny of modern science. Patting these aside, 
there remains in his so-called uncharacterized sorts — an immense 
number that have been seen by his predecessors ; and I think 
that when such authorities as Cuvier, Valenciennes, Richardson, 
Sleeker, Kaup, Euppell, &c., admit them as distinct species, after 
having studied them, they are at least as much entitled to be 
believed as the zoologist who has not even seen them. 
Taken, for example, the sorts brought back by myself from 
the central parts of South America, and deposited at the Garden 
of Plants of Paris, we find that Dr. Gunther considers many of 
them as identified with species of Cuvier and Valenciennes, 
when, in the Ichthyological part of my Travels, I give them 
as distinct. Perhaps the imperfection of my descriptions 
may have led him to believe in their identity ; but it must 
be remembered that those specimens were all compared with 
Cuvier and Valenciennes’ types, aside of which they are 
placed in the Museum, and that this examination was not 
only done by myself, but in many cases by Messrs. Valen- 
ciennes and Dumeril, and in all cases by Mr. Guichenot. On 
the other hand Dr. Gunther appears not to have examined the 
Parisian collection, which is certainly the most important in the 
world, on account of the immense quantity of typical specimens 
it contains. 
I also think that Dr. Gunther carries too far the modern 
tendency of uniting sorts that were considered as distinct, and, 
misled by this principle, he has formed a certain number of 
artificial species which do not exist in nature. It is well known 
that in the class of fishes colours are, in general, subject to such 
alterations as not to aftbrd, as a general rule, specific characters, 
as they do in almost all the other divisions of the animal king- 
dom, and that these characters must be looked for in the forms 
