24 
ORLAND E. WHITE 
apply the term catacorolla. The data on each are given in some detail, 
followed by a short discussion and summary. 
The work was done in the Laboratory of Genetics, Bussey Institu- 
tion of Harvard University, under the direction of Prof. E. M. East, for 
whose kindly interest and criticism, I wish to express my appreciation. 
The material was obtained from various pure line cultures of 
Nicotiana species, which had been under observation for several 
years. All pure species used in this study bred comparatively true 
and no abnormal variations appeared in them, except in Nicotiana 
langsdorffii grandiflora, which was subject to petalody, and gave 
evidence of being a hybrid as it was heterozygous for yellow and blue 
pollen, the true form according to Comes (1899) having only blue 
pollen. 
I . Petalody 
This teratological character is an extremely common feature of 
garden flowers, and, as usually found, is variable even among the 
stamens of the same flower, i. e., one stamen may possess it, or it may 
be present in two, three, four or all of them. On one stamen, the 
petal-like outgrowth from the filament, which constitutes the char- 
acter, may be very small, while another filament in the same flower 
may show an anomalous enlargement from three to ten or twelve 
times as great. It presents its extreme form in the common double- 
flowered races of Dianthus, Rosa, Prunus and Ranunculus. The 
majority of gardeners as well as many scientists believe that such 
double-flowered races can be created from single-flowered varieties by 
selection. A very excellent treatment and historical resume of this 
subject is given by de Vries (1906, Chap. 17) in which he produces 
historical proof that many of our common double-flowered races arose 
suddenly and in full possession of their peculiar character. His 
experimental studies led him to assign doubleness because of its 
variability, to the category of "ever sporting" characters. In many 
of our cultivated races, double-flowered plants quite faithfully repro- 
duce themselves if they are fertile at all. The majority of these 
races have arisen as mutations, the causal factors of which are largely 
unknown. Among horticulturists the belief is prevalent that intense 
cultivation is responsible for the anomaly, but there are no data from 
controlled experiments to support such a belief. Peyritsch (Goebel 
1900, I, p. 195) induced all degrees of doubling in the floral organs of 
