No. 388.] FOUR CATEGORIES OF SPECIES. 291 
and to locate a particular point in a continuous series of this 
kind is manifestly a different problem from that of finding the 
island which is its extremity. The necessities of the first case 
are met by observing the appearance, development, and perhaps 
the decline of some important structure, such as a tooth. The 
line of natural succession is the important fact, and the classifi- 
cation, as far as there is any, is here of necessity artificial, 
touching at only one point, that of living forms, which lies in 
a distinct plane. It is only when considered with reference to 
their own horizons that fossil species are distinctly compa- 
rable to the specific islands of to-day; and to fail to consider 
the breaks in the horizontal series, because the vertical series is 
theoretically continuous, is one of the more important instances 
where a confusing element has been unnecessarily and illogi- 
cally introduced. It is one task of taxonomy to recognize the 
facts of the present; and this need not be complicated, but 
should rather be assisted, by a knowledge of what has been or 
a suspicion of what may yet be. Nor is it necessary to be 
frightened by the captious warning that the species now exist- 
ing will give place to something different to-morrow. The 
“now ” is at least sufficiently extensive for our purposes. As 
to the coming change, posterity will not know its rate nor 
direction unless we certify the details of the present reality. 
The careful description and preservation of types, instead of 
works of supererogation, are of the utmost importance, not as 
means of specific limitation, but as giving fixed points about 
which accretions of knowledge may gather. The recognition 
and adequate designation of natural species or higher groups 
are frequently matters of extreme difficulty, requiring extended 
and careful study which the agnostic naturalist jauntily avoids, 
sometimes by professing interest in “ more important” prob- 
lems. To tell from a few specimens whether a new form rep- 
resents a distinct species or not is frequently impossible, and 
opinion must be provisionally based on the analogy of better 
known relatives ; but this initial difficulty in no way affects the 
practicability of the subsequent settlement of the question by 
more extended observation. ; 
Whether the segregation of a new group has been accom- 
