Hazel-ivood with Reference to Conductivity of Water . 567 
and does not ascend with curve D. The widest vessels observed were 
in H8 and Hio, in the lowest internodes; the maximum diameters for H7 
and H9 are smaller. With this absence of very wide vessels may be con¬ 
nected the initial ascent in curve E for H7 and H9, which contrasts with 
the initial descent in curve E for H8 and Hio. 
Curve F is drawn upon another sheet for each specimen. It indicates 
the total number of conducting elements estimated for the complete trans¬ 
verse section of the wood at the selected internodes. This curve shows 
a general decline, with more or less irregularity towards the apex. At the 
internode corresponding with the maximum in curve C, the wood is seen 
from curve F to be particularly rich in vessels. 
Upon the same sheet is drawn also a set of graphs to show the numbers 
of conducting elements of different diameters, which make up the total 
numbers given in curve F ; the diameters are given at intervals of 4 /jl 
as they were measured. On the whole, in any one transverse section, 
the narrow elements are more numerous. The internodes proportionately 
richest in wide vessels are those towards the middle of the shoot. 
Owing to insufficient data, no figures for lengths of vessels are given 
in this paper, but I have reason to think that such data would not alter 
materially the conclusions arrived at from transverse characters. 
Conclusion. 
In this paper I have described a statistical method of investigating the 
constitution of wood from the standpoint of its efficiency for conducting 
water. It is an attempt to record definitely the number, size, and distribu¬ 
tion of the elements in the wood which are concerned in this process, and to 
present the results in a graphical form. The method is intended to serve 
as a basis for correlating with anatomy the facts of specific conductivity 
obtained by experiment, and possibly to bring to light further facts in this 
connexion. 
In stool shoots of the hazel the figures have shown a very consider¬ 
able variation in the constitution of the wood formed during the first 
season. On the whole there is a general decline in total conductivity 
and a general rise in specific conductivity from the base of the shoot to its 
apex. The explanation, in this case, is simple : it has been shown to be re¬ 
lated to the provision for the other functions carried out by the wood, and 
chiefly to the greater proportion of mechanical elements in the lower part of 
the shoot, providing the support necessary in this region. 
I wish to acknowledge in this place my indebtedness to Professor 
Farmer for the help he has given me in this work, and to thank him 
for his interest and encouragement. 
