1904] COPELAND—VARIATION OF CALIFORNIA PLANTS 409 
tecognize it.2 Only the two starred in the table were collected after 
my attention was attracted to them by their many differences; they 
grew with interlacing branches in open ground by a roadside. No 
well-shaded bushes were chosen, nor shaded branches on the bush. 
While the immediate environment of the leaves was thus eliminated 
as a factor in making them different, the condition of the roots was 
left uncontrolled. I have no doubt that docking its roots would 
cause one of these shrubs to produce thicker, rounder, and more 
entire leaves, and I believe that these would tend to be pubescent 
beneath and rolled inward. But no visible differences in the ground 
were associated with the different variations. That the correlated 
Variations are not common functions of any outside agent is evidenced 
oo by their measure of independence—flat pubescent, entire green 
leaves, etc. The data in the table do not approach the range of 
Vanation of the individual leaves nor indicate the frequency with 
Which the usually correlated variations are otherwise combined. 
The frequent correlation of thickness of leaf with pubescence, rolling 
backward, and evenness of margin might be due to such a com- 
bination of characters in heredity as many recent writers on inheri- 
can assume; but I would rather ascribe it to mechanical factors 
Vhich operate as the plant grows. 
The Tange of the variability of this Rhamnus is not more remark- 
able than the absence of any well-defined type from which variation 
‘aa ded as taking place; or, to express it in the usual eh = 
oa cgi the curve representing any one of the varying char 
__s Under discussion would be conspicuously broad and low, 
ae a well marked maximum and steep slopes. That this 7 
: a. the shape of the leaf the tabulated irae os ge oe 
ordinat, were plotted with abscissae of 0.2, beginning wit se 4 
€S would be 1, 2, §, 6--S, 3%, oe ee the length were p otte 
th * curve with abscissae of 1°" (4o per cent. of the shortest leaf), 
ea would be 3, 4, 6, 6, 5, 2, —1- lf a curve = es 
a the thickness, with abscissae of 20 units of the sphe 
F 40 per cent. of the thinnest leaf), the ordinates would be 2, 
“Thave one branch of whose leaves the ratio of average length to average = 
: 435; iced i in 
5*1-00, but as I noticed its slender leaves before collecting it is not included 
the table, 
