No. 388.] FOUR CATEGORIES OF SPECIES. 293 
necessity of recognizing species and genera created by the 
gardener and poultry fancier. Taking as a criterion of amount 
the minute differences by which some distinctly segregated 
species, such as those of Antennaria, are distinguished, the 
impartial botanist of conceptual proclivities would find himself 
more than occupied in providing for the scientific recognition of 
horticultural novelties. With degree of difference as the cri- 
terion of classification, the annual increase in the species of the 
genus Rosa would be very considerable, for the divergences 
resulting from crosses and selection are often as great as those 
existing between wild species of undoubted validity. 
The four types of species may then be enumerated as follows: 
I. Lhe phylogenetic species, a division or section of a line of 
biological succession. 
2. The insular or segregated species, the living end of a line 
of the preceding category. 
3. The incipient species, preferably known as the subspecies, 
which is a subdivision of the second category, being a group of 
individuals showing distinctive characters and a tendency to 
segregation, but still connected with other groups by normally 
existing intermediate forms. 
4. The artificial species, the result of man’s interference in 
nature, by which the specific islands of the second category 
have been, as it were, remodeled or connected by causeways 
or bridges, the natural tendency to isolation and segregation 
having been reversed through human agency. 
In one case only does nature limit the species; in the other 
three the species are arbitrary in the sense that their bounda- 
ries are formal and to that extent conceptual. It were much 
better if other terms could be made available, so that the desig- 
nation “species” might be reserved for its original use with 
the second category. For the third the appropriate term “ sub- 
species ” is increasing in favor, while to the fourth there seems 
to be no good reason why the popular designation “ variety ” 
should not-be restricted. Even here it is not the absolute 
degree or amount of difference which determines the desirability 
of independent recognition for a new form, but the constancy 
and utility of the differences. It is fortunately still true that 
