34 
JOUKNAL OF THE EOYAL HORTICULTUEAL SOCIETY. 
senting the relative values of the crops in comparative experiments. The 
method is, obviously, not free from objection, but it is the least objection- 
able which we have at present been able to devise. One evident defect in 
it is that it takes no account of the colour and quality of the fruit. 
In the case of trees, however, we have hardly arrived at the stage yet 
when the crops are of sufficient importance to call for great attention, 
and growth is the paramount consideration at present. We have been 
obHged to consider, therefore, how the growth may best be measured. 
This would seem to be almost as simple a matter as appraising the crops, 
till one comes to study the question carefully. The height and average 
spread of the trees have been determined annually, but these have not led 
to any very satisfactory results, and fail altogether in cases where the 
systems of pruning are varied. The diameters of the stems at a certain 
height from the ground are still less satisfactory, for the measurements 
are too much influenced by the irregularities of the stems and the 
inequalities of the ground. The number and weight of the twigs removed 
at the annual prunings have also been recorded, but they afford still less 
reliable data, since the removal of growth depends not only on the length 
of new wood formed, but on the position of the branch in the tree. The 
most satisfactory means of measurement, no doubt, is to ascertain the 
total increase in weight of the tree by weighing it before planting, and 
lifting it and weighing it again after the completion of the experiment. 
This plan we have adopted wherever practicable, but its application is 
evidently very limited ; it can only be adopted where the experiment is 
complete within two or three years, and even then would not apply in 
all cases, such as in a comparison between pruned and unpruned trees. 
The method which we have finally adopted, and which appears to be of 
more general application, is to determine the weight of the leaves. It is, 
of course, a matter of common as well as of scientific knowledge that the 
vigour of a tree is dependent on, and, therefore, may probably be measured 
by, the leaf area which it exposes. The leaves are the lungs of the tree, 
and the seat of the manufacture of starch and cellular tissue through the 
absorption of carbon dioxide and solar energy ; and the number, size, 
texture and colour of the leaves will always give us roughly a fair indi- 
cation of the health of a tree. Accurate measurement, however, is not 
an easy matter. The total leaf area of the tree is, doubtless, the proper 
thing to measure, but, although we have determined this roughly in 
several cases, the labour which such a determination would entail with 
the trees, now that they are of a fair size, would be altogether prohibitive. 
We have therefore adopted the plan of collecting a certain number of 
leaves (6 or 10) from each tree, taking in every case the sixth leaf from 
the end of a branch, and determining either the area or the weight of 
these. Recently we have relied on the weights, taking care, however, that 
the leaves are properly dried before weighing, and that the weighings, in 
the case of any set of experiments, are all made under similar conditions. 
Of course, such a method is obviously open to many sources of doubt 
and error, and it is not at all apparent that it will yield trustworthy 
results ; it is only by a critical examination of the results obtained in 
similar experiments on different varieties, and of these results when 
compared with those obtained by other methods, that we can feel justi- 
