in existence. When such variations arise in naturally inbred 
plants, such as peas, they generally breed true at once; but in 
cross-fertilized species, such as corn, purification by selection 
must be practiced to secure a true breeding strain. 
Inasmuch as most of the higher plants and animals result from 
a union of two microscopic cells, and as each of these cells may 
carry the hereditary units which together with environment are 
responsible for the characters of these organisms, it follows that 
every sexually produced animal or plant that breeds true for 
several successive generations must have at least two of these 
units, identical in every respect, for each of the characters in which 
it breeds true. The only exceptions are those cases where a 
number of characters result from one kind of unit, in which case 
two of these units, exactly alike, may be responsible for several 
characters, as, for example, in the characters pink flower color, 
red leaf axil color, and gray seed coat color in peas. 
Granting tor the time being that the Mendelian method of 
analysis is applicable to practically all the characters of the world 
of living organisms, — that plants and animals are made up of in- 
dependently inherited units, which, in combination with environ- 
ment and with each other, express themselves as characters; and 
granting further that these units are unvariable and not modifiable 
by selection, how can we arrange and account for all the variation 
we see about us ? 
If the conceptions of Mendel and his followers hold as regards 
the composition of organisms, variations may only occur: — 
1. When the complex of heredity units constituting 
the organism remains the same, and the external environ- 
ment ( e.g.y soil, climate, food) is changed. 
2. When one or more of these units is added to or sub- 
tracted from the organism (through crossing), the exter- 
nal environment remaining unchanged. 
3. When both the complex of hereditary units and the 
environment are altered. 
4. When the environment remains the same, but the 
complex of hereditary units (the organism) is altered by 
the “addition” of a new hereditary unit, or the “subtrac- 
tion” of an old one (mutation). 
All four of these statements find illustration in almost every 
plant breeder’s experimental garden. At the Brooklyn Botanic 
Garden, the plant material for the study of heredity is largely 
made up of peas, castor beans, and corn, in which many units 
have been isolated. Studies are also being made on variation in 
the Boston fern and its varieties. In Boston, at the Bussey Insti- 
tution of Harvard University, the material is principally tobacco 
and corn; and among animals, mice, guinea pigs, rats and rabbits 
1 5 
