An Onagraceous Stem without Internodes. 51 
that this colony of naturalized plants contains three general groups 
of forms: (1) 0. Lamarckiana, which is almost if not quite 
indistinguishable from the race of De Vries’ experiments and 
certain of De Vries’ mutants, (2) a series of plants which are the 
same as O. grandiflora Ait. in cultures from one of its original 
localities, in Alabama, and (3) a considerable number of races 
having unexpected characters which are in many cases not inter¬ 
mediate between the O. Lamarckiana and O. grandiflora races. 
From a further analysis of this population of inter-breeding 
individuals, in comparison with other series of cultures of QEnothera 
races from various parts of the world, and by a comparison of their 
hereditary behaviour with that of related races whose origin is 
known, one may hope to learn something further regarding the 
manner of origin of these forms. And more particularly, it should 
be possible by such means to analyze the De Vriesian evolutionary 
factor of mutation, and discover in how far it is merely a process 
of hybrid splitting and in how far (as certain results show) it is a 
more deep-seated germinal disturbance, resulting in part from the 
effects of previous crossing, from the direct influence of changed 
environment, or from some internal and unknown cause. 
Experiments dealing with these and related questions are still 
in progress, but the results are far too extensive to attempt to deal 
with in this communication, the purpose of which is merely to 
describe a peculiar development which occurred in one of my 
cultures of these Lancashire plants several years ago. A note on 
this subject was read at the Winnipeg meeting of the British 
Association' in 1909. The culture in question was grown entirely 
in the tropical greenhouse of the University of Chicago, the seeds 
being planted in July, 1907. Under these conditions of high 
temperature and high humidity, the plants belonging to the O. 
grandiflora series of forms remained rosettes until May, 1908, when 
they all formed shoots, and flowered during the following summer. 
But the 28 plants belonging to the O. Lamarckiana series of forms 
(which is more strongly biennial in habit than 0. grandiflora) 
nearly all remained rosettes throughout the experiment, which 
lasted nearly two years, or until May, 1909. During this time, 
however, growth was continuous, new cycles of rosette leaves being 
continually added above while the older ones died away below. In 
this way a short stem was produced, without internodes and covered 
with leaf-bases in quincuncial arrangement. Plate II shows such 
1 Rept. Brit. Ass., Winnipeg meeting, p. 677, London, 1910. 
