THE GIRTH-INCREASE OF TREES. 81 
also for annual observations, selecting as far as possible such as 
had smooth bark and regularly cylindrical stems. These having 
become ineligible as a connected set in 1892—although some of 
them became available afterwards, as they recovered at various 
dates from transplantation—were replaced by the final young 
set of twenty trees, available from 1892 to 1895, but then dis- 
abled for my purpose by a severe pruning. 
The plan followed in dealing with the monthly observations 
is to give zw extenso the results for the set of 1892-95, not 
hitherto published, to compare these with the results yielded by 
the other young set of 1887-91, and to make use of the original 
set of older trees, 1882-87, only incidentally. 
In considering the records of 1892-95, I shall first try to give 
the aggregate results, and then describe the conduct of each 
species separately, inquiring at the same time how far the results 
may agree with those obtained from other young trees observed 
in 1887-91, in sofar as the same species-happen to have been 
examined in both these periods. 
In the first place, however, it is well to state that neither the 
number of trees of each species, nor the length of time during 
which they have been under observation, is sufficient to warrant 
the deduction of precise conclusions or definite laws. At first 
sight, indeed, it might seem that three specimens of a species, of 
similar ages, growing in the same locality, and under observation, 
two of them from 1887 to 1891, and the other from 1892 to 1895, 
should yield true averages, but that this is not so is proved by 
the occasional quite contradictory conduct of a tree in one year 
as compared with the other three or four years, or what is still 
more striking, by the contradictory conduct of one tree during 
the whole four or five years of observation, when compared with 
the other two trees. Neither is it always easy to account for 
this erratic conduct, although among probable reasons may be 
suggested—temporary unhealthiness, not, it may be, betrayed 
by theappearance of the tree; individuality of character, as 
when we see two trees of the same species, and equally vigorous, 
of which one invariably comes into leaf much earlier than the 
other ; difference of age, which, even when slight, has, I suspect, 
considerable influence in early youth; difference of position, 
which even within narrow limits may place trees under very 
