4 SMITH: STUDIES IN THE GENUS LUPINUS 
tative parts of the plant merits reasonably close attention. Evi- 
dently the environment, and especially competition with plants of 
other genera, has much to do with whether the plant is low or tall. 
Presumably the amount of available plant food may determine 
- whether the plant is acaulescent and unbranched, or more or less’ 
caulescent and branched. The axial terminal bud, unless injured 
by some external cause, may be depended upon to produce a 
fruitful peduncle. Branching is, of course, merely the result 
of the development of axillary buds, and it is common, in large 
well-branched plants, to find matured or ripened seeds in the pods 
of this axial peduncle when the racemes of at least some of the 
branches are still in bloom. A low unbranched acaulescent plant 
is a record of its own environment, and a dry barren hillside or 
desert area the habitat in which it is able to both exist and persist. 
On the other hand, a tall plant, branched more or less above the 
base perhaps, almost always comes from an environment also 
favorable to plants of other genera; and abundance of nourishment, 
assisted by the stimulation of competition, is fundamentally re- 
sponsible for the extensive vegetative growth developed. The 
“‘fistulous”’ character, especially, is certainly the evidence of a 
rapid growth during a warm moist season. I see no “‘specific’”’ 
characters in the stem structures of these plants, excepting, to 
some extent, in L. luteolus Kell.; but I have admitted into my 
treatment of varieties more than one definite reference to such 
stem characters, especially when coupled with other characters or 
supported by geographical distribution. The length of the pet- 
ioles and the size of the leaflets and stipules vary with the stem 
portions and seem to me to be of no value for specific distinctions. 
Even the pubescence failed to give satisfaction in my effort to 
draw specific lines. 
Most reliable of all, I have concluded, is the position of the 
flowers in anthesis and soon after anthesis. This may not always 
be strictly constant in all the flowers of a particular raceme, nor 
in all the racemes of a certain plant; but I am convinced that, with 
occasional allowance here for some little variation, this character 
will prove to be as positive a one as can be found in the Micro- 
carpi, and certainly is more usable than the very indefinite dis- 
tinctions as yet proposed by those who have treated any of the 
