BRAINERD: FOUR HYBRIDS OF VIOLA PEDATIFIDA 259 
nations in fours, of these eight pure elementary characters, are 
to be found in these 41 plants. 
As showing how a hybrid tends in successive generations to 
eliminate its hybrid characters, ever becoming simpler and finally 
pure, we note that starting with the F, plants, necessarily hybrid 
in all the opposed characters of the two parent species, we have in 
the next generation only two such hybrids, and in the 171 plants 
of the 2d generation, none; indeed, the law of probability calls 
here for only one in every 256 offspring. 
_ My first recognition of V. pedatifida X sororia was in a package 
of living plants sent May 22, 1907, by Dr. H. V. Ogden of Mil- 
waukee, collected at Upper Nemahbin Lake, Wis., growing with 
both parents, and considered by him as ‘‘ doubtless a hybrid.”’ 
The same thing, however, had been sent me some three years 
earlier by Dr. Greene as a specimen of his V. Bernardi, collected 
by himself at Dixon, IIl., June 18, 1898. In Leaflets 1: 184. 
Ja 1906, the plant is transferred to V. perpensa, then first de- 
scribed. I have recently examined the three other specimens 
there cited and regard them all as forms of V. pedatifida X 
sororia.* Also V. fallacissima Greene, Leaflets 1: 185. Ja 1906, is 
another form of the same hybrid from western Missouri— Bush 141, 
Lee’s Summit, Mo., July 8, 1899. Other specimens are: E. J. 
Palmer 3345 and 3393, Webb City, Mo., April 19 and May 5, 
1911; Mary O. Pollard 6, Yorkville, Ill., May 16, 1909; L. M. Um- 
bach, prairies, Clarendon Hills, Ill., June 21, 1899—distributed in 
1911 from United States National Herbarium as “‘ V. palmata.”’ 
4. Viola nephrophylla X pedatifida hyb. nov. 
V. Wilmattae Pollard, Proc. Biol. Soc. Wash. 15:178. Au 1902T 
Foliage much as in V. papilionacea X pedatifida, nearly 
glabrous, palmatifid with several narrow lateral lobes; corolla 
‘deep violet, 2 cm. broad ”’; petals markedly villous and sepals 
with slightly scarious margins, as in V. nephrophylla. 
It has not been practicable to secure living plants for cultures. 
But the alleged parent species were both growing in the cafion, 
* I therefore wish to correct my statement in Bull. Torrey Club 37: 584. 1910 
that V. perpensa Greene is the western form of V. palmata—a too hasty inference from 
the fact that a V. palmata specimen in my herbarium from Aurora, Ill., was 
named V. Bernardi by Mr. C. L. Pollard. : 
pe is 404,924 United States National Herbarium, Mrs. Wilmatte P. 
{ The ty 
Cockerell, Sapello Cafion (c. 8000 ft. alt.), Beulah, New Mex., May 5, 1901. 
