228 
HARLEY HARRIS BARTLETT 
in the European garden form shows that it has been hybridized with 
some broad-petaled species, and cannot therefore be considered typical 
of Oe. cruciata, de Vries ^ has named it Oe. cruciata var. varia. He used 
this nominal variety for the crosses between ''Oe. cruciata'' and other 
species which are described in the second volume of Die Mutations- 
theorie. Because of its unknown, mixed ancestry it was an undesirable 
type for use for further experiments. De Vries, therefore, wrote to this 
country for seeds of the true Oe. cruciata, which he had no reason to 
suspect was not well known here, for Small^ had only recently (1896) 
restored it to specific rank, under the name Onagra cruciata, and had 
cited it not only from Massachusetts but also from Vermont. The 
Vermont record was probably based upon specimens collected by 
Grout at Brattleboro and Vernon; the localities were later published 
in the Flora of Vermont."^ 
In response to his request, de Vries received seeds and rosettes 
from MacDougal, collected at Sandy Hill, New York, near Lake 
George, and seeds from Robinson, collected at Jaffrey, New Hamp- 
shire — all supposedly Oe. cruciata. These were grown at Amsterdam 
in 1903. The material from Sandy Hill included two very distinct 
types, both of which were found among the plants from seeds as well 
as among those from rosettes. The differences between the types 
were slight, but striking and absolutely without transition. Most of 
the individuals had rather broadly linear petals and comparatively 
thick flower-buds. The rest had narrowly linear petals and slender 
flower-buds. Robinson's seeds from Jaffrey gave a third type, di- 
verging from both of the others in its much longer calyx tube and more 
slender stem, inflorescences and foliage. De Vries^ interpreted his 
findings as follows: "It seems not improbable that Oenothera cruciata 
includes a group of lesser unities, and may prove to comprise a swarm 
of elementary species, while the original strain might even now be in a 
condition of mutability." In regard to the two types from Sandy 
Hill, he said: "Probably two elementary species were intermixed 
^ De Vries, H. On Atavistic Variation in Oenothera cruciata. Bull. Torr. Bot. 
Club 30: 75-82. 1903. Die Mutationstheorie. 2: 100. 1903. 
s Small, J. K. Oenothera and its Segregates. Bull. Torr. Bot. Club 23: 169. 
1896. 
^ Brainerd, Jones and Eggleston. Flora of Vermont, 1900, p. 63. 
8 De Vries, H. Species and Varieties, 1905, p. 590. See also letter quoted by 
MacDougal, Vail, ShuU and Small, Mutants and Hybrids of the Oenotheras, 1905, 
p. 12, 
