DUNN — UNIT CHARACTER VARIATION IN RODENTS 
127 
species of plants and animals. There has resulted from this work the 
elaboration of the chromosome theory of heredity, for the details of 
which, and the supporting evidence, the interested reader must be 
referred to the original works, especially as summarized in two publi- 
cations of Morgan and his co-workers (1915 and 1919). 
For our purpose it is sufficient to note in brief that the theory supposes 
that the differential representatives of heritable characters are located 
in the nuclear material of the egg and sperm cells, more precisely in 
those remarkably constant and individual organizations of chromatin 
known as chromosomes, which appear at the time of cell division and 
which probably retain their individuahty even in the resting stages 
of the nucleus. Of the evidence it must be observed that the interpre- 
tation and proof of the theory rest entirely on the study of linkage, 
or associated inheritance. This phenomenon is observed in the ten- 
dency which characters exhibit of remaining through several genera- 
tions in their original combinations, resulting in an alteration of the 
expected Mendelian ratios based on independent assortment. This 
tendency may be absolute, in which case linkage is said to be complete. 
More often it is partial, that is, characters originally associated may 
separate in a certain proportion of instances, or characters originally 
separate may become associated. This change in the relationships 
of genes is known as “crossing-over’’ and it provides a quantitative 
measure of the strength of the tendency toward association. In terms 
of the chromosome hypothesis it is interpreted as an interchange of 
parts and of the genes which the parts carry, between two members 
of a chromosome pair, so that two genes originally resident in one 
chromosome may come to lie in two chromosomes and may thence be 
distributed to separate gametes and exhibit their effects (unit characters) 
in different individuals. For any two characters the number of times 
crossing-over occurs is found to have a characteristic value and this 
value is stated as the percentage of times crossing-over occurs as 
evidenced by the frequency of individuals possessing the two characters 
in the new combination. One other importent aspect of these measura- 
ble breaks in hnkage is that from the linkage strength may be inferred 
the proportional distance apart of hnked genes. From cytological 
evidence crossing-over is supposed to take place between homologous 
chromosomes in the hybrid at the time when these chromosomes are 
intimately twisted one about the other. Breaks resulting in a separa- 
tion of characters are then supposed on mere physical grounds to be 
more frequent between genes located far apart in the chromosomes 
