284 L. Agassiz on the Primitive Diversity 
Fossil Mammalia are comparatively too well known to call 
for many remarks, after what has been said above. Let us 
only remember that the number of fossil species found in 
Brazil alone equals the whole number of Mammalia known 
to live at present in that country; that the fossil Mammalia 
of New Holland compare already favourably with the living 
species of that continent; and that the locality of Mont- 
martre alone has yielded as many large Mammalia as occur 
all over Europe; and the Mauvaises Terres in Nebraska as 
many as may be found in North America now. So that, if 
we grant simply that among vertebrata the diversity has been 
increasing with the successive introduction of their different 
classes, the number and diversity of these different classes 
at each period has been as great as it is at present. 
These facts are of the utmost importance with reference 
to the great question of the order of succession and gradation 
of animals in the different geological periods. They cut 
away for ever one of the arguments upon which the asserters 
of the development theory have insisted most emphatically. 
Before it could be granted that the great variety of types 
which occur at any later periods has arisen from a successive 
differentiation of a few still earlier types, it should be shown, 
that in reality in former periods the types are fewer and less 
diversified ; and we have now shown that this is so far from 
being the case, that in many instances the reverse is really 
true. I have already attempted elsewhere to show in out- 
lines what is the real order of succession of the great 
types ef the animal kingdom, I need not therefore re- 
peat here what may be gathered from the diagram in the 
Zoological Text Book I have published jointly with Dr Gould. 
I shall limit myself to a few more general remarks upon the 
special difficulties involved in a more thorough investigation 
of the subject. 
The study of the order of succession and gradation of the 
organized beings which have inhabited our globe at different 
periods, presents indeed difficulties of more than one kind. 
Unhappily these difficulties have seldom been all considered 
in their natural connection by those who have ventured to 
consider the subject in its whole extent; thus presenting 
