846 _. The American Naturalist. [October, 
have undoubtedly been inherited, as in the familiar facts of 
atavism, and they must effectually prevent the detection of 
the initialstages in the development of any new characters 
under investigation, which may in fact have been potentially 
transmitted for a number of generations. 
Reversion, prepotency, and the influence of a previous im- 
pregnation, are conspicuous obstaclesin the way of tracing the 
immediate, or incipient indications of the inheritance of any 
particular acquired character which it may be desirable to 
perpetuate by judicious selections. 
In many of the arguments relating to the heredity of 
acquired characters it appears to be tacitly assumed that each 
particular character is transmitted as an entity, regardless of 
its interdependent relations to other parts of the organism, 
and especially with the specific functional adaptations of the 
organs of nutrition which have made its development possi- 
ble. 
As pointed out in a paper read in Section F, at the Rochester 
meeting of the Am. Ass. Adv. Sci., physiological changes in the 
organism must precede any manifest modification of structural 
characters, and the transmission of a morphological peculiar- 
ity must, therefore, involve the transmission of the functional 
activities through which it has its origin. It was also shown 
that a habit or bias of the nutritive processes in a certain 
direction, may be transmitted for a number of generations 
without any visible morphological evidence of its existence, 
and that, in thelapse of time, it may lead to the development 
of obvious structural changes that are recognized as new 
characters. Experimental methods in biology are too crude 
to admit of a recognition of these preliminary steps in the 
development of new characters, which must be taken into 
account in making a satisfactory verification of any of the 
processes of evolution. 
The artificial conditions to which our domesticated plants 
and animals are subjected, intensifies their susceptibility to 
variation, and there appears to be a constant tendency to re- 
version when any unfavorable conditions prevail in their 
treatment. Underordinary management, repeated systematic 
