A Genetic Study of Plant Height in Phaseolus Vulgaris. 59 
The modification would then also have to be used in interpret- 
ing the results secured from crosses between two bush races or 
two pole races of different heights, for there seems to be no 
fundamental difference between such crosses and crosses between 
a pole bean and a bush bean. But, for the results from crosses of 
unlike bush or pole beans, the multiple-factor hypothesis affords 
a much more simple and direct explanation. It accounts for the 
intermediate development in Fi, for the wide range of variation 
in Fo, and for the constancy of some and inconstancy of other F2 
types. 
In many respects the results secured from crosses between 
pole and bush beans resemble the results obtained from crosses 
between hooded and Irish rats (MacCurdy and Castle 1907), and 
hooded and wild rats (Castle 1912). When hooded rats — char- 
acterized by pigmented head, shoulders, and forelegs and median 
dorsal stripe and white over the rest of the body — were crossed 
with Irish rats — characterized by pigmented sides and dorsal 
surface and a variable white area on the ventral surface — the 
Irish pattern was dominant in Fi and segregation into Irish 
and hooded patterns with a 3-1 ratio occurred in Fo, but the 
pigmented area in the hooded segregates was increased as was 
also the range of variation in amount of pigmentation. From this 
MacCurdy and Castle conclude: 
"Again, though the inheritance is clearly Mendelian, when hooded and 
Irish rats are crossed, the gametes formed by cross-breds are not pure, but 
modified, each pattern being changed somewhat in the direction of that pattern 
with which it was associated in the cross-bred parent." 
A strain of hooded rats with extensive pigmentation and a 
strain with restricted pigmentation were established by selection. 
When crossed with wild rats — totally pigmented — each hooded 
strain behaved as a simple Mendelian recessive, but the hooded 
segregates from the cross of wild rats with the extensively pig- 
mented hooded pattern were more extensively pigmented than 
the hooded segregates from the other cross. Crosses between 
hooded rats with restricted pigmentation and hooded rats with 
extensive pigmentation gave results similar to most crosses where 
the parents differ quantitatively. (Castle and Phillips 1914.) 
The fact of simple Mendelian segregation in such crosses has 
led Castle to maintain that only a single unit-character is involved 
in the experiment (Castle 1912, 1914), tho he admits the possibility 
that additional factors may be concerned (Castle and Phillips 
1914) as first suggested by East (1912) and later discussed by 
the Hagedoorns (1914) and Muller (1914). That the hooded 
pattern differs from the Irish pattern in a single genetic factor 
will be readily admitted. That there are no additional factors 
