174 
Journal of Agricultural Research 
Vol. XI, No. 4 
but in some cases still more data are needed, in view of the fact that the 
linked groups may be less closely bound together in Pisum than in 
Drosophila, because of difference in chromosome dimensions. In several 
cases, involving small totals, the approximation between the classes 
actually obtained and those theoretically expected is not very close. As 
soon as more data are obtainable probable errors for these ratios will be 
calculated. In Table III, the ratios actually obtained are given first and 
the calculated expectation directly beneath. Each factorial combination 
is separated by a line and the combinations themselves (somatic) are 
expressed by descriptive character symbols, explanations of which are 
given in a note below the table. Factors (Tl) and R are used inter¬ 
changeably, as they are closely linked. The totals represent the F 3 
populations from crosses of different varieties, though each of the two 
varieties involved in a single cross differed in the designated factors. 
The data from the individual crosses with the numbered designations of 
the varieties involved, are given in the appendix. Each group of crosses 
involving one type of factorial difference is designated so as to correspond 
with the designation in Table III, as exempli gratia (i), (2), etc. All data 
except those involving group 7 are from the Brooklyn Botanic Garden 
cultures. The data for group 7 are the combined results from the studies 
of Vilmorin (1911), Bateson (1909), Bateson et al. (1905), Pellew (1913), 
as well as from the writer’s own studies. 
The employment of data from crosses involving many different varie¬ 
ties to demonstrate the independent inheritance of specific factors, such 
as those given in Table III, as compared with data from crosses involving 
only two varieties with the desired factorial differences, is open to some 
criticism on the hypothesis of multiple factors. Nilsson-Ehle, (1908, 
1909), East (1910), Morgan et al. (1915), Shull (1914), and many others 
have secured F 2 ratios involving only one character difference, which, 
however, gives a di-, tri-, or poly-hybrid ratio. 
Apparently the same characters in plants and animals of the same 
species in other experiments have given an ordinary monohybrid 3 :1 
or 1 :2 :1 F 2 ratio. These results (dark-brown glume color in oats, 
yellow endosperm color in maize, pink eye color in Drosophila, capsule- 
shape in shepherd’s-purse {Bursa bursa-pastoris ), etc., are interpreted 
by their discoverers as showing the presence of two (in case of a 
dihybrid ratio) or more factors, each of which by itself gives rise to prac¬ 
tically the same effect, expression or character, as when both are con¬ 
cerned. Further, as the ratio indicates, they are inherited independently 
of each other, and according to Morgan, Sturtevant, and their colleagues 
(1915), they are located in different chromosomes. By the method in 
use in the present study, different factors in different varieties giving the 
same or very nearly the same “somatic” expression could not be dis¬ 
tinguished; hence, one can not be certain that he is always experimenting 
with the same factor. 
