53 
these successive groups is purely arbitrary. The division of 
classes into orders, and these into sub-orders and genera, is 
highly convenient, and as already noticed, on a general view, 
not without strong warrant of facts in its favour ; but in its 
precise limitations it is arbitrary. Naturalists are perpetually 
divided not only as to which are species and which varieties, 
but as to where genera begin and end, how far orders and 
sub-orders are to be distinguished, and especially under what 
head particular species or genera are to be ranked. The 
constantly increasing divergence that appears as we ascend 
the scale almost necessitates such intermediate groups being 
introduced, and yet the gradations are in many cases so fine, 
the connecting links so numerous, as to render it a difficult if 
not a hopeless task to define and arrange these groups in a per- 
fectly natural manner. Again, precisely what might have been 
expected if all these successive groups were the irregularly 
divergent but yet related descendants of a single progenitor. 
Once more, it is to be noted that the differences which 
distinguish these various grades of groups from one another 
vary exceedingly as to the organs and characters which they 
concern. Now it is the most important which are found 
to differ, now the least; nor does this variation accord in 
any way with the importance of the classificatory distinction. 
Thus we have some orders of plants (as Oruciferse) where 
the number and position of the stamens, the arrangement 
of the petals, &c., are alike throughout ; the generic and 
specific characters being obtained for the most part from 
organs of less importance. And again, we have other orders 
(as Connaracese), where the most radical characters are found 
to vary between genus and genus ; or in some cases even 
between species and species. Had the contrary been the 
case, and the most fundamental organs afforded the charac- 
teristics of the larger groups, the less fundamental those of 
the subordinate ones, and so on in regular gradation, — had 
this been so, the arrangement and relations of living’ beings 
would have presented a symmetry and manifest method 
strongly suggestive of especial design and arbitrary plan. 
The opposite to this, however, — irregularity, ununiformity, 
apparent lawlessness, — was naturally to be expected, if all 
these groups were really the diversified offspring of a common 
parent, since such diversification would be certain to proceed 
irregularly in different directions. And exactly thus we 
find it. 
We come now to another test. If the Darwinian hypothesis 
be true, then not only have large groups of species descended 
from single progenitors, but the mode of descent has been by 
