16 BULLETIN 905, U. S. DEPARTMENT OF AGRICULTURE. 
that these contributions are composed of units which are handed on 
generation after generation without change. The union of sperm 
cell and ovum in its bearing on heredity may be compared to the 
mingling of two collections of solid beads instead of to a mixing of 
two liquids. Certain characteristics, such as coat color, depend on 
such a small number of these units for their development that the 
separate ones can easily be identified, given symbols, and followed 
from generation to generation. Most characteristics, including size 
and conformation, depend on such a large number of units for devel- 
opment that the effects of the separate ones can not easily be dis- 
tinguished. The inheritance is naturally more or less of the blend- 
ing type, but a large number of phenomena, such as prepotency and 
the effects of inbreeding and crossbreeding, can be understood best 
by the theory that the hereditary basis is composed of a limited 
though fairly large number of unchanging units. 
The statement that these units are unchanging applies to ordi- 
nary experience. It must not be taken too literally, however, as 
if this were true there could be no progress. Cases have been clearly 
established hi which a unit must have become modified so that its 
effect is changed or, in many cases, apparently wholly lost. Differ- 
ences between individuals depend on the possession of different 
alternative forms of certain of the hereditary units. 
It may be added that the experimental evidence indicates that 
there is in general equal inheritance from sire and dam with respect 
to all kinds of characters. With the exception of a rather unusual 
class of cases, which is discussed later, it appears that the sperm cell 
contains a full set of the hereditary units characteristic of the kind 
of animal and that the same is true of the egg cell. The fertilized 
egg cell thus receives a double set of the units. 
COLOR AND ALBINISM IN GUINEA PIGS. 
A concrete illustration will bring out the behavior of these heredi- 
tary units in a typical case. It has been mentioned already that 
albino guinea pigs always breed true. Stocks of colored guinea pigs 
can also be obtained which breed true in the sense that they never 
produce albinos. The first cross between such stocks results wholly 
in colored young. Color, therefore, is said to dominate over albi- 
nism, and, conversely, albinism is said to be recessive to color. If 
these crossbred young, whether male or female, are bred back to 
the albino stock, it will be found that only 50 per cent of the young 
are colored, the other half being albinos. These albinos, when 
crossed with each other, produce only albinos, and this is true of 
their descendants indefinitely. The power to produce color seems 
to have completely dropped out of their make-up. If their colored 
brothers or sisters are crossed with the albino stock, the result is, as 
