A quantitative Study of Variation, Natural and 
Induced, in pure Lines of Silene Noctiflora’). 
Eugene P. Humbert, M.S. in Agr. 
The subject matter of this paper is based upon a study, through 
three generations, of a pure line of Silene noctiflora. For the stati- 
stical data about seven thousand five hundred plants have been 
measured for the following characters: height of plant, width of plant, 
number of branches per plant, and number of seed pods (capsules) 
per plant. These characters were chosen not only because they 
include some of the important features such as size, form, habit of 
growth, and productivity, but they are such as may be quickly and 
accurately measured. The commonly used statistical constants are 
employed in dealing with all measured characters. 
These seven thousand five hundred plants were under experimental 
control and were handled and grouped so as to give data on four 
more or less separate and distinct problems which are discussed 
separately in this paper. They are: 
I. The influence of certain chemical injections on variation. 
II. Effectiveness of selection in pure lines. 
III. Bud variation. 
IV. Influence of food supply on variation. 
Along with the statistical work mentioned above there were some 
fifteen thousand plants under careful observation throughout one 
season for the detection of any plants showing marked deviation 
(mutation) from the normal fluctuating variability of the species. 
It does not seem necessary or advisable, in this paper to take 
up a discussion of the prevalence of variation, or its importance in 
1) Contribution IX. Laboratory of Experimental Plant Breeding, Cornell Uni- 
vertisy. 
Presented to the University Faculty of Cornell University for the degree of 
Doctor of Philosophy. 
Induktive Abstammungs- und Vererbungslehre. IV. II 
LIBRA 
NEW Y! 
BOTANI 
GARD! 
