219 NATURAL AFFINITIES—INFERENCES. 
tion, or, on the other hand, that they will continually 
degenerate if those controlling influences are withheld. 
Now what is the evidence of the facts which regularly 
come under our observation? We find that ever since 
sorghum and imphee have been cultivated within the 
United States, they have not only freely commingled when 
grown in each other’s vicinity, but they have also freely 
amalgamated with doura, or Guinea corn, and with broom 
corn, and new and fertile varieties may thus readily be pro- 
duced and perpetuated. No other grasses exhibit an ex- 
ample of such prolificacy in the hybrids of well-determined 
species. } | 
But if they are distinct species, there must be found some 
characters to fasten upon which can be regarded as constant, 
and such also as demand recognition, as belonging to spe- 
cies. Such a character is certainly not presented in the 
saccharine quality of the juice, for the same is possessed in 
a greater or less degree by other kindred species of grasses ; 
nor can any others properly be so regarded. 
Among the most decisive tests of specific identity are the 
following: two races may be regarded as specifically dis- 
tinct when there are no intermediate gradations tending 
to connect them, and, conversely, two races may be re- 
garded as specifically identical when they are connected 
together by close intermediate gradations. (Carpenter. ) 
Applying this principle, we discover, in collections 
embracing the different kinds of this cane grown in the 
United States, that all the prominent characters, which 
might be thought at first sight to indicate specific diversity, 
are manifested in a greater or less degree by all alike, and 
that there is almost every possible gradation between ex- 
treme forms. Chinese sorghum, the different kinds of 
imphee, doura corn, and broom corn, are distinguishable 
from each other chiefly by such trivial characters as the 
