( 140 ) 
similar manner. As the petiole is not in the same plane with 
the leaf-surface it was necessary to take great care in press- 
ing down the leaflets so as not to distort their position or alter 
their length. 
After all the leaves of the two series of one species had 
been plotted in the above manner, the measurements could be 
made at leisure after drawing the lines from the dots to the 
center. The positions are measured in degrees in a clock- 
B : W C ge K 
0 
K V\g ql 2 
vy. wl < , W 
I i 
\ if y Y 
T- w aS" go q 
La S= ri 
ayy W 
w 
A & 
(One half natural size.) arthenoctssus quinguefolia. Ten 
leaves of operated series as plotted on page of note-book. 
wise (+) direction from the line representing the petiole. The 
lengths are in millimeters and include the distance from the 
tip of the leaflet to the center of the leaf. Thus we have in- 
cluded in the length of the leaflet, the blade and its very 
short petiole. 
The special difficulties in each of the species may be men- 
tioned in a few words. In Parthenocissus the leaves taper 
veryjgradually near the tip and have a tendency to dry up 
and to curvein this region. In the case of a very sharp curve 
at the end the prolongation of the main portion of the blade 
wasjtaken instead of the curved tip as it stood. Ina few 
cases the drying up of the tip made it necessary to approx- 
w i 
