1895.] The Philosophy of Flower Seasons. 99 
In the solution of the problem of the flower seasons of a 
given flora I think that the period of no plant should enter 
as a factor, if it is so far removed geographically that at its 
blooming time it does not become a competitor of any plant 
of that flora. Any number of flowers adapted to the same 
conditions may bloom at the same time if they are so widely 
separated that they do not interfere with one another, but it 
would be an obvious disadvantage for very many of them to 
bloom at the same time in the same locality. In the latter 
case a separation of the blooming times would be advantage- 
ous. On the other hand, there are some objections to the use 
of data derived from a local flora’, though I think they are not 
so serious. Phenomena which seem to find an explanation in 
a limited field may in fact find their true explanation in con- 
ditions outside of that field. Even in the case of a local flora 
the time of blooming is likely to be indicated as too long, 
since it is based upon the early dates of early seasons, and 
the late dates of late ones. Such data give rise to error by 
making it appear that the period of an early species overlaps 
with that of a later one when in fact the two species never 
have flowers in bloom at the same time in any season. It is 
hardly practicable to avoid this, since observations confined 
to a single season are liable to be too fragmentary. 
To note that a given family of plants is highly specialized 
and that it agrees with Mr. Clarke’s generalization by reaching 
its maximum in summer, does not help one to understand either 
the general position of the family or the blooming time of a 
single species, and the difficulty remains the same whether the 
species blooms before or after the bulk of the family, or whether 
its season coincides with the maximum. The main fault that 
may be found with his elucidation of the subject is that it is 
implied that the general principle of the late blooming of 
highly specialized flowers is an explanation of the blooming 
phenomena; for, whenever a flower agrees with the general- 
ization, it is left as if it were thus explained, while, if it is an 
exception, its period is accounted for under the considerations 
which we have numbered. And it must have been a striking 
fact to the readers of the paper that the exceptions yielded so 
