BRAINERD: FOUR HYBRIDS OF VIOLA PEDATIFIDA = 255 
tifida, already discussed in this paper. Of these no. 1950 flowers 
freely in May with showy flowers, that often appear also in July 
and August; but apetalous flowers are rare, and neither sort has 
been found to produce seed. The three. others are also nearly 
sterile, bearing only 5-10 seeds to a capsule; but none of the 
three turns out to be like the Pollard plant from Yorkville in being 
a first cross (or F,), the form to be selected as the starting point 
for experiments on the laws of inheritance in hybrid offspring. 
No. 1952 has an uncut leaf as in V. papilionacea and a green 
colored capsule as in V. pedatifida, both recessive or reversionary 
characters never found in a first cross. No. 1949, on the other 
hand, has the hybrid leaf and the hybrid capsule color but the 
buff seeds of V. pedatifida and is therefore another subhybrid. 
No. 1947 consists of five plants, the seeds of which differ in color, 
and the leaves of which, though somewhat incised, display at 
least three unlike patterns; the five plants therefore must be 
considered the offspring of an earlier hybrid. 
The remaining plants of the 1909 collection are equally variant 
formas but of V. pedatifida X sororia. No. 1957 is of dwarf habit 
and has the compromise or hybrid leaf; but though vigorous and 
multiplied by division into eleven plants, it has failed to yield a 
single seed. No. 1955 (six plants by division) has a leaf more 
deeply cut that the others, and this style of leaf reappears in all 
its offspring. In this we see a reversion, far from complete but 
stable, to the leaf of V. pedatifida. No. 1953 (again six plants 
by division) has the pure green capsules of V. pedatifida, as have 
also its offspring, and so is another subhybrid. But the two 
remaining numbers, 1954 and 1948, seem to be the desired first 
product of hybridism, all the four pairs of opposed characters in 
the double parentage appearing in a compromise form in both 
numbers. A flowering specimen of 1954 was distributed in my 
“ Violets of eastern North America, 1910,” no. 121; and in no. 122 
are shown two sister offspring, one with the uncut leaves of V. 
sororia, the other with the parted leaves of V. pedatifida.* No. 
* In one of our large herbaria, where the work of mounting is done by novices, 
these two offspring were considered too unlike to appear on the secs sheet. : 
the plant with uncut leaves was mounted over ticket 122; while the sister plant with 
parted leaves was placed on the sheet with no. 121, which indeed it more closely 
resembled. That two plants so dissimilar should come from one self-fertilized 
parent has seemed incredible even to certain “ . 
