512 
Chibnall.—Diurnal Variations in the 
will be lost. Now the inward and outward flow of nitrogen through the 
petiole cannot be tapped, consequently it is impossible, by direct chemical 
analysis, to determine the nature of the nitrogenous substances that pass out 
of the leaf. But if it can be definitely established that the total nitrogen in 
a leaf falls at night, then, by chemical analysis, it may be possible to deter¬ 
mine from what substances (such as protein) the outgoing nitrogen has origi¬ 
nated. This would be an important advance, for though it is known that 
the protein is synthesized from nitrates, &c., translocated from the root 
system, very little is known as to its function in the leaf cell and the com¬ 
position and ultimate fate of its degradation products. Such analyses have 
already been made (Kosutany (6), Suzuki ( 12 ), Schulze and Schiitz ( 11 ), 
Chibnall ( 2 )), but it is not proposed to discuss them here. Their interpre¬ 
tations depend primarily on the method by means of which the comparison 
of the day and night leaves have been made. In the following pages these 
methods are discussed in some detail, and the conclusion is drawn that one 
of them, that which estimates the nitrogen as a percentage of the dry weight 
of the leaf, gives an inaccurate and misleading result. 
In comparing the day and night samples of leaves three methods have 
been used to estimate the total nitrogen present, namely, those in which it 
is expressed as 
A. Weight in terms of a certain number of leaves, 
B. A percentage of the dry leaf-weight, 
C. A percentage of the fresh leaf-weight. 
It is proposed to examine each of these in detail to ascertain if the 
diurnal change given by them is reliable. 
Method A. If every leaf on a plant was directly comparable with every 
other, this method would give the absolute diurnal change ; but since this 
is not so the question of sampling errors at once arises. The two workers 
whose results obtained by this method are discussed later have not deter¬ 
mined these. An idea as to their magnitude, in the case of annuals, may be 
gathered from some experiments made by the present author with the 
leaves of Vicia Faha maj'. Twelve plants, about six weeks after they had 
appeared above ground, were used. All the leaves from each plant were 
picked and weighed. Probable errors, calculated by Peter’s formula, are 
given in Table I. 
Table I. 
Shoiving Probable Errors of Leaves from Vicia Faba maj. 
Number of 
Weight of 
Average 
leaves per 
leaves per 
weight of 
plant. 
plant in 
a leaf in 
grm. 
grin. 
Mean of 12 plants 
35 
1 975 
0-574 
Probable error of the mean of 12 plants 
± i -17 
±o -99 
+ 0-02 5 
Percentage error of the mean 
3-34 
5 -o 
4.42 
