ZOOLOGY. 229 
sylvania.” * The I., II., and III. propositions are prefixed as pre- 
liminary : — 
I. That genera form series indicated by successional differences 
of structural character, so that one extreme of such series is very 
different from the other, by the regular addition or subtraction of 
characters, step by step. 
I. That one extreme of such series is a more anes 
type, nearly approaching in characters the corresponding extreme 
of other series. 
III. That the other extreme of such series is excessively modi- 
fied and specialized, and so diverging from all other forms as to 
admit of no type of form beyond it. 
VI. That therefore the differences between genera of the same 
natural series are only in those characters which characterize the 
extreme of that series. 
For the highest groups in the animal kingdom we must accept 
the definition of Cuvier, Von Baer, and Agassiz, for the present, 
that they are primary, because they represent different primary 
plans of structure. For the lowest grade of groups (genera) 
the definition above given (Prop. VI.) will be found to represent 
groups to which the definition given by Agassiz f will also apply ; 
viz. that “their special distinction (¢. e., of genera) rests upon the 
ultimate details of their structure.” I believe that the definitions 
given by Agassiz to the three intervening grades of divisions — 
viz. of families, orders, and classes — are far nearer a representa- 
tion of nature than any other ever given. They are as follows: — 
Classes are defined “by the manner in which the plan of the 
branch is executed; Orders, by the degrees of complication of 
that class-structure; amilies, by their form as determined by 
structure.” Natural science is under great obligations to Pro- 
fessor Agassiz in this, as in other points. 
These definitions are, however, better perceived after the groups 
are constituted, but in practice are not sufficiently exact to serve 
as the crucial test in the cases which may arise. The simple 
method indicated in-our propositions above will, it appears to us, 
serve to solve many of the more difficult questions which arise 
during the attempt to state the true relations of organic beings. 
We may now apply these principles to the groups of the class 
* Trans. Am. Philos. Soc., 1866, p. 397. 
{ Contrib. Nat. Hist: U. S.,i. pp. 163, 170. 
