176 
Journal of Agricultural Research 
Vol. XI, No. 4 
Two such multiple factors as A and A ly each capable of producing pink 
flower color, each located in a separate chromosome, and each inde¬ 
pendently inherited from a third such factor having the same expres¬ 
sion, could bring about considerable confusion in such an attempt as the 
present to determine the factorial groups of Pisum. With data from 
crosses of two varieties having seven or more factorial differences, no 
such difficulties would be encountered, and perhaps even with the method 
used above the chance of encountering more than one such factor, is 
comparatively rare. As soon as two true breeding varieties differing in 
seven or more suitable factors can be secured from crossing experiments 
the independence in inheritance of the factors listed in Table III will 
again be tested. 
From among 16,384 or more F 2 progeny of such a cross, if each factor 
proved again to be independently inherited, there should be over 128 
(2 7 ) visibly distinct forms, representing 2,187 distinct factorial combina¬ 
tions. Provided the cross involved 8 factor differences, each inherited 
independently of the other, 256 distinct forms would be theoretically 
expected and if a cross were made involving 25 such factor differences, 
each independently inherited, the F 2 generation, if it were practicable to 
grow a large enough population (4 25 ), should consist of at least (2 s5 ) 
33,554,432 visibly distinct forms. The classification of such a popula¬ 
tion, even when obtained, would be impracticable, if not impossible. 
It is brought into the present discussion to emphasize the enormous 
amount of variation possible through crossing, for even on the chromo¬ 
some-crossing-over-linkage hypothesis of Morgan and his coworkers 
(1915), plants and animals are known with more than enough factor 
differences and chromosomes to have 25 independently inherited groups. 
For example, Nicotiana tabacum, the commercial tobacco, as observed by 
the writer (1913), has 24 chromosome pairs and the recent lists of Tischler 
(1916) and Ishikawa (1916) cite plants with a much higher number. 
As pointed out by East (1915), if the facts and theories regarding inher¬ 
itance in Drosphila spp. are found to be true for organisms in general, it 
is very important that plant breeders should know the number of chromo¬ 
some pairs in the plants with which they are experimenting. When one 
contrasts the number of forms that can be derived from such crosses as 
mentioned above with the number of recognized plant species (approxi¬ 
mately 225,000) and considers that to the plant breeder and horticul¬ 
turist many of these forms are as distinct as taxonomic species, one is not 
surprised at all that a certain well-known taxonomist and student of 
phylogeny, Lotsy (19), should advocate that all evolutionary change of 
a hereditary nature is due to crossing. 
Other factors in addition to those listed in Table III have been tested 
out with each other, but only part of the possible combinations with those 
given in Table III have been made, so these data are reserved for a later 
paper. 
