558 Holmes.—A Study in the Anatomy of 
sections were made at the centre of each of a number of particular inter¬ 
nodes chosen at various heights along the shoot. Drawings of these sections 
were made on millimetre squared paper, by means of the camera lucida, and 
the total areas of the section as a whole and of the wood in the section were 
calculated. Then, in a typical section of the wood, the area of which was 
determined in a similar manner, the diameter of the cavity of every water¬ 
conducting element was measured by means of an eyepiece micrometer, and 
recorded. From these observations a number of statistical results were 
arrived at for each section in the following order: First, an average figure 
was obtained for the diameters of the elements in question ; and secondly, 
the number of these elements in a unit area was worked out. The results 
of these calculations were used to determine the proportion per cent, of the 
area of the wood occupied by the cavities of the vessels and tracheides ; and 
from this percentage figure and the figure for the total area of the wood it 
was possible to calculate the total area occupied by these cavities in the 
complete section. The same observations were available also for working 
out the proportion of wide to narrow elements in the section. 
For the purpose of obtaining information as to the variation in the 
lengths of the vessels it was necessary to apply a method of a different type. 
It is not possible to observe the lengths of these elements in longitudinal 
section, for they extend through several centimetres. I had the opportunity, 
however, of making use of Professor Farmer’s apparatus designed for the 
injection of vessels. The material used for this purpose consisted of fresh 
stool shoots collected in February. The shoot was cut into a number of 
suitable lengths for injection, the distances of the several injection surfaces 
from the base being recorded. After injection, transverse sections were cut 
at measured distances from each injection surface to determine the maximum 
distance reached by a specially prepared colloidal solution of Indian ink. 
Accuracy. Naturally the figures obtained for the diameters and 
numbers of the elements measured are approximate only. In transverse 
section the water-conducting elements are very variable in shape, and most 
of them are by no means circular; the areas were worked out for circles, 
however, and as far as possible the diameter recorded for each element was 
that which would give the nearest area with the appropriate formula in 
view, the measurements being made with a §-inch objective. All the 
measurements and calculations were made in a uniform way so that the 
figures obtained should have a comparative value. 
Results. The data obtained by these means are presented as far 
as possible graphically, in the form of curves. Taking first the characters 
shown in transverse section, I have drawn for each shoot a set of graphs to 
show the change in the constitution of the wood from base to apex. In 
constructing the graphs the numbers of the internodes were marked along 
the base line, taking an equal interval for each internode, irrespective of its 
