290 THE AMERICAN NATURALIST. [VoL XXXIII,. 
shallow, and tortuous channels, like the Florida Keys. These 
last are, indeed, a cartographer’s nightmare comparable to the 
species of Aster and Sphagnum. But whether the natural- 
ist’s theories help or hinder, his descriptive task will not be 
complete until the facts of nature have been recognized and 
recorded, until all the islands have been located and mapped. 
That this work is in a confused and backward state is partly 
due to the conceptual fallacy of species and genera, which has 
led us to attempt to map the islands without securely locating 
them by the designation and preservation of types. 
To divide a continuous series of forms into “genera” and 
“species ” is a thoroughly artificial process analogous to the 
marking of feet and inches on a measuring rod, and is one of 
the instances where the naturalist may be said to “make” the 
“ species.” The specific islands of living organisms are, how- 
ever, not made, but discovered by the systematist, and located, 
not primarily with regard to their disposition in an ancestral 
series, but with reference to other separate groups. Whether 
a specific island is large or small, circular or irregular in shape, 
is not subject to our determination; and the naturalist who 
expects to apply throughout nature a general concept of species 
based on the “size” or “weight” of characters is about as 
well equipped for his work as a map-maker who should under- 
take to draw a chart of the Florida Keys without other tools 
than a circular die. With such an implement he could in a 
general way indicate the location of a few well- separated 
islands like the Canaries, but the other task is Eoo 
impossible. 
Although chronologically the first, the problem of systematic 
paleontology, on account of the fragmentary nature of its 
material, must depend largely upon the results gained in the 
study of existing forms. Isolated fossils may be classified, 
for the sake of convenience, in so-called species and genera ina 
manner analogous to that used with living types ; but when any 
approximation to a complete phylogenetic series can be made 
out, that method is no longer logically applicable. The chasms 
which separate species in the present do not interrupt the un- 
broken line of ancestry which stands below every existing type, 
