338 Gates. 
progressive sort, where each member of a series exhibits a condition 
intermediate between the members coming before and after it. As 
instances of this in Oenothera may be mentioned, the progressive 
decrease in size of the stem-leaves from below upwards, a similar 
decrease in the thickness of the stem, and in the length of side 
branches. In O. rudbrinervis the stem-leaves also show a decrease 
in crinkling from the bottom of the stem upwards, as well as a pro- 
gressive change in size and shape. Similarly there is a decrease in 
the amount of pubescence on the stem towards the top. 
In the case of the flowers, there is a well-known progressive 
decrease in size after the earlier flowers have bloomed; the style 
frequently becomes shorter, though not always in proportion to the 
decrease in size of the petals. The mean number of stigma lobes has 
also been shown (SCHOUTEN, 1908, p. 120) to drop gradually to the 
modal number, four. All these things and others show a tendency to 
a gradual tapering off in the vitality of the plant during ontogeny. 
Certain other variations, however, do not appear to be of this 
sort. For instance, in the study of variability in the red color pattern 
of the sepals in Oenothera, there is no apparent decrease in the modal 
condition of an individual during the season. And in the ontogeny 
of the rosettes, as will be shown in greater detail elsewhere, there is 
not a single progressive series of changes in successive leaves, but 
several distinct types of leaves usually appear. 
I. Variation in the red color pattern on the sepals 
of Oenothera. 
A. O. rubrinervis. 
I shall first present the data of color variation in O. rubrinervts 
in which it has been studied most extensively. O. rubrinervis differs 
from O. Lamarckiana in the following characters (Mutationstheorie 
I, 235—238). 
1. Smaller leaves with red nerves and gray tomentose surface, 
much less crinkled than in O. Lamarckiana. The whitish gray color 
of the leaves DE VRIES states to be partly due to increased hairiness, 
but more largely the result of a light effect, due to the arched surface 
of the epidermal cells without hairs, this arching being insignificant 
in O. Lamarckiana. The difference in size and crinkling of rosette 
leaves is purely relative, and a sufficiently large culture produces 
