94 BRAINERD: HYBRIDS OF THE PALMATA GROUP 
were grown in 1907: one quite glabrous, in this reverting to 
V. Stoneana; the other three more or less pubescent. From each 
of the four plants offspring were obtained in 1909. Those from 
the glabrous plant were also all glabrous; two of the pubescent 
plants bore both pubescent and glabrous offspring; the remaining 
pubescent plant bore nine offspring, all pubescent, thus indicating 
that in this instance the pubescence of the mother plant was a 
stable character. This happens to be exactly the relative number 
of these three forms of offspring required, on the average, by the 
laws of Mendel. 
Viola latiuscula X triloba hyb. nov. 
Leaves moderately 3-5-lobed, sparsely pubescent on the 
petioles and veins of the young leaves; capsules about 8 mm. long, 
maturing 10 or 12 brown seeds; progeny heterogeneous. 
For my acquaintance with this hybrid I am indebted to the 
discernment and diligence of Mr. B. H. Slavin of the Park Depart- 
ment of the City of Rochester, N. Y. On June 19, 1909, he sent 
me from Salamanca, N. Y., six odd-looking plants, all glabrous; 
four with leaves somewhat lobed, subcordate-ovate, blunt- 
pointed; two with leaves uncut. As they flowered and matured 
the following season, I detected certain marks of V. latiuscula: 
a crimson tinge in the early spring foliage, and a granular rough- 
ness along the upper edges of the petiole. At my request Mr. 
Slavin revisited the station for further collections July 6, 1910, 
and sent me in the autumn an excellent suite of the various forms 
to be seen in that colony. For a better apprehension of their 
relation to each other, I borrow some of the symbolism of Mendel. 
For brevity he uses letters instead of phrases, somewhat in this 
fashion: 
Let A=cut-leaved Let B = pubescent 
“a= with leaves uncut ** b= glabrous 
* Aa = with leaves somewhat uncut ** Bb = somewhat pubescent 
All of these six characters are found in the above described hybrid 
or in the parent species; in the offspring of the hybrid they should 
be redistributed in all possible combinations. It is evident that 
each one of the three characters in the first group may combine 
with each of the three in the second group, making in all nine 
different combinations. These are given in the following table, 
