2 BULLETIN 1468, IT. S. DEPARTMENT OF AGRICULTURE 
the transmission of characters, in an effort to remove some of the 
uncertainties of breeding and to provide a basis for a more orderly 
procedure for future improvement. 
The manner in which new characters arise is still in dispute, but 
with the progress of investigation it has become increasingly clear 
that the process of the transmission of characters is largely mechani- 
cal. For purposes of convenience the characters manifested to our 
perceptions are thought of as resulting from the interaction of genes, 
invisible determiners located in the germ cells. These hereditary 
elements are transmitted from one generation to another ordinarily 
without change. Some characters are relatively simple, requiring 
for expression but a single gene, but all degrees of complexity exist, 
some of which exceed the limits of our present ability to analyze the 
hereditary constituents into their component parts. 
The genes occur in groups; groups which in some organisms cor- 
respond to the haploid number of chromosomes, leading to the belief 
that the chromosomes are packages of genes. 
Experimental evidence of exceptional accuracy has been acquired 
from many sources showing that the genes are arranged in a linear 
series with respect to one another and always in the same order. 
Further, it has been possible to measure the distance separating 
the genes in any given group, not in actual units of linear measure 
but relatively to one another. The distance between related genes 
expresses their correlation and measures the extent to which they will 
appear together in inheritance. Genes which appear together in 
inheritance are spoken of as linked and the phenomenon is known 
as linkage. 
To study linkage relations the procedure is to cross two individuals 
that differ with respect to two characters, the genes of which are 
located in the same chromosome. In the first generation the domi- 
nant form of both characters is expressed. The first-generation 
individuals will produce gametes of four kinds representing the four 
possible combinations as follows: (1) Both characters like the 
female parent; (2) both characters like the male parent; (3) one 
character like the female parent, the other character like the male 
parent; (4) like 3 with the characters reversed. The nonparental 
combinations, 3 and 4, are called the crossover class. The percent- 
age of the total number of gametes that falls into this class is the 
percentage or rate of crossing over. 
If the second generation is the result of back crossing the first- 
generation individuals with individuals homozygous recessive for 
both of the characters, the characters of each effective gamete pro- 
duced by the hybrid individual will be expressed in the offspring. 
If the two characters are completely independent, the four combina- 
tions will occur in equal numbers and the two combinations repre- 
senting the crossover class will be 50 per cent of the total. If 
crossing-over percentages in excess of 50 were found they would 
represent disherences, or cases where the characters descended from 
opposite parents tend to become associated in inheritance. 
With the demonstration of the linear arrangement of genes sepa- 
rated by measurable distances, the question has arisen as to the sta- 
bility of these distances. This question of stability is of more than 
theoretical interest, since plant improvement through breeding often 
depends upon the facility with which character combinations can be 
