interest to apply to these Boston fern sports, and others, not 
pertinent to this account, the test of actual conditions, to deter- 
mine whether they are fit to survive on a natural selective basis 
as well as on that of artificial selection. 
An interesting and illuminating differentiation between the 
natural and artificial selective methods may be postulated in the 
case of the development of the races of domesticated and wild 
canine forms. Some savage— more than once it must have hap- 
pened — finding a den of wolves, made captives of the young as 
playthings. Sometime, among such a captured litter, there must 
have occurred at least one puppy more susceptible of taming, 
which liked captivity, and came to depend on its human captor 
for protection. Many wolf puppies have been captured, and more 
or less tamed, but generally less, because it is a hereditary char- 
acteristic that the average wild animals shall distrust confinement, 
and resist taming. The rare variation, the tameable wolf puppy, 
would delight the man captor, and in time it might add its value 
as protector of the household, warning off its former associates. 
But what would be the fate of such a milder, less ferocious puppy 
under wild conditions ? To ask is to answer. Nature does not 
seem to have selected wolves on the basis of mildness, and tame- 
ableness. 
To apply the conclusion of the foregoing : from the Boston 
fern there have developed variations which answer the test of 
specific differences as far as distinctiveness is concerned. Whether 
they would all survive under natural conditions remains to be 
tested. Some variations, however, have been found, and have 
occurred at the Brooklyn Botanic Garden, which answer the 
difference test, and would undoubtedly survive under the con- 
ditions of natural selection. 
But the Boston fern has gone much further -along the lines of 
demonstrating evolution than merely producing six distinctive 
new forms. Each of these has continued the process. While re- 
maining constant in their own characteristics, following the law 
of heredity in the great mass of progeny, each of the six cited 
above has given rise to new forms, well distinguished from 
their parents, and like them able to breed true, i.e., the}' are 
variations of the kind called mutations. Two illustrations will 
suffice. From the Pierson fern, itself twice divided, has come the 
Barrows fern, somewhat smaller and more consistently twice 
divided. From the Barrows fern has come the Whitman fern, 
three times divided. (See Plate II). From the Whitman fern has 
come the Smith fern, four times divided, and lacy; and from this, 
an apparent end to this line of progression, the five times pinnate 
Craig fern. Similarly, from the Giatras fern has come another 
type accentuating the reduced size, a dwarf variety less than one 
sixth the size of the original Boston fern, named, as it happened, 
the New Jersey fern. (See Plate I.) From the original ruffled 
forms, have come new forms with a greater degree of ruffling; 
from divided varieties there have appeared dwarf divided forms. 
It appears that these types of variation may be shuffled and 
6 
