410 BOTANICAL GAZETTE [DECEMBER 
I, 3, 6, 6, 3, 3, 3; or if the abscissae represent 4o units, beginning 
with 40, the ordinates become 3, 9, 9, 6. The shape curves of varia- 
tion of this plant must be clear from these three illustrations. Of 
course, the curve could be made steeper and narrower by increasing 
the value of the abscissal units; but by this process, if carried to an 
extreme, any curve can be reduced to a vertical line; and 4o per 
cent. of the shortest or thinnest leaf ought to be a large enough unit. 
The curves for the entireness of the margin and the pubescence of 
the lower surface would be of the same shape as these if there were 
scales by which they could be measured. These numbers are too 
few to make symmetrical curves, but as an objection to the validity 
of the conclusions this is largely removed by the fact that each num- 
ber is an average of a considerable number of measurements. 
ARCTOSTAPHYLOS TOMENTOSA Lindl. 
This is the common manzanita of this region. Some authors 
have detected more than one species in what I include in it, but I 
am sure nobody would do so on my material. It was all collect 
in a limited region near the top of the mountain south of Woodside, 
using the same care to preclude the influence of local differences of 
environment as in collecting Rhamnus. ‘Variable as it is, the mam 
zanita has proved a much less favorable subject for this work than 
the Rhamnus, for several reasons. The most variable feature is the 
pubescence of both stems and leaves, and I have not found it feasible 
to measure this, because the chaff varies in size as well as in abun- 
dance, and is irregularly deciduous. Another difficulty is that the 
leaves formed at the ends of the seasons are smaller, narrower, and 
more entire than those typical of the plants; this made it necessary 
to discard some of the leaves, and to decide more or less arbitrarily 
where the typical leaves ceased on each axis. As this left too few 
leaves on each axis to furnish a safe average, I used all the fit “ee 
on the twigs of a small branch of each shrub. Tables I — a 
give the averages from thirteen plants, and the details of Gece ei 
vidual leaves of one plant, to show their individual variability ne 
data under “‘margin” show the average number of teeth on each ips 
of the leaf, and the average per cent. of the distance from base 2 
apex at which the most apical tooth is found. Under base and ape 
their angles are given. 
