1058 The American Naturalist. [December, 
ations, he finally succeeded in producing forms that contained 
minute spherical refringent bodies, which in some cases exter- 
nally resembled true endospores, but biologically differed from 
them materially as they were destroyed by heating for 
two hours at 60? C while the normal endospores are 
among the most resistent bodies known. This is a well 
authenticated instance, where the morphological character of 
reproduction has been modified, while the salient physiologi- 
cal features of the germ remained constant. This change had 
evidently been brought about through the influence of exterior 
conditions, and so deeply had the inner constitution of the germ 
been affected that it transmitted the character to its progeny 
although the normal conditions of development did not favor 
its production. 
ere we to admit the evidence of physiological variation, 
we would find abundant proof among the bacteria that this 
group of organisms were more or less profoundly modified in 
their functional chgracters. Physiological variations have so 
far received but slight attention at the hands of biologists, but 
in bacteriology the truth of De Varigny’s words that species 
must be defined not only by means of their anatomical char- 
acters, but also in terms of physiological differences, has been 
amply confirmed. Bacteriology has been forced to add physio- 
logical to the morphological diagnosis in the study of these 
minute forms. 
That there is a variation in the characters of certain forms 
. in a perfectly natural state is readily seen in the case of some 
contagious diseases like cholera. Itis known that the peculiar 
individualities of the cultivations isolated from the different 
epidemics are so marked, that an expert can tell at a glance 
whether the culture in question descended from the germ 
found in the Naples outbreak, or in Egypt, or from India. As 
one epidemic is often the result of the transference of germs 
from another, and the two germs therefore more or less related, 
it is reasonable to infer that environment has much to do 
with their modification. But we are not confined to the evi- 
dence of physiological variation as afforded by examples 
under natural conditions. These conditions are too fluctuat- 
