27 
If we had all the forms of life that have ever existed before us, 
it might be urged that the differences in every direction were so 
gradual that the species might all be considered as belonging to one 
genus ; nay, we may go further, and suggest that we should probably 
have to admit that all life, all individuals, were of one species. But 
even granting this, the close resemblances between the individuals 
comprising the various groups, and also the differences between the 
individuals of one group and those of all other groups, would still 
exist, and in spite of the fact that none of these groups would be 
suddenly or abruptly terminated on their limits, yet the groups— 
generic, tribal, sub-family, and family—themselves would exist. 
I need not labour the point further, that all our studies prove to 
us that more or less homogeneous groups of species exist, and that 
this being so, the recognition of genera is necessary. ‘This being 
conceded, it is clear that as soon as the abstract idea of what con- 
stitutes a new genus is formed, the genus must be defined by a name. 
This name will then convey to the mind two fixed ideas, viz. (1) the 
assemblage of species included therein ; (2) its own relationship to 
allied groups. 
The groups that have been formed at various times and from 
different causes sometimes agree very exactly with what we con- 
sider to be convenient ; hence we speak of ‘a well-defined genus,” 
meaning thereby one in which the species, which should have inter- 
graded with other genera, have from some cause or other become 
extinct. On the other hand, the gaps are sometimes so well filled, 
7.e. in reality have never been formed, that a large group may show 
no positions at which we can divide it into genera without separating 
somewhat closely allied species, and yet the extreme species are much 
wider apart than the distance that ordinarily satisfies us for different 
genera. Groups of this kind, in which a large number of species 
which merge into each other, but which have at their extremes 
species with different characters, thereby showing the independent 
development each has undergone, should be carefully studied in order 
that the subsidiary evolutionary centres may be discovered, and the 
group thus divided into its natural sub-genera, for it is very incon- 
venient to have large unwieldy genera in faunistic work. 
Here, then, are the two lines on which all our generic work should 
be based :—(1) The attempt to separate species into natural groups. 
(2) The subdivision, if possible, of a natural group containing a large 
number of species into two or more subsidiary but still natural and 
evolutionary groups for convenience. It must not be forgotten that 
the natural types of these subsidiary groups are usually quite as far 
removed from each other as are those of more clearly defined genera 
around which the extinction of intergrades has taken place. 
We can illustrate these two points with the Acronyctas, about the 
subsidiary groups of which there has recently been so much dis- 
cussion, We had better, perhaps, do so in Dr. Chapman’s own 
words. He writes: “I tell you that the British Acronyctas split 
