IN PLANTS HAVING IRREGULAR COROLLAS. 389 
patible with constant and perpetual usefulness. Tt is clear that 
the degree to which this difficulty applies to any case is propor- 
tional to the complexity, perfection, and singularity of the con- 
trivance. 
In addition to the foregoing objection, a further difficulty 
arises when we try to figure to ourselves the kind of transitional 
stages by which the evolution of a complex mechanism may have 
been brought about. "We are here met by an entire want of 
evidence as to the nature of such changes, and it is not easy 
even to conceive any hypothetical plan on which they may have 
occurred. 
On the other hand, the objections to supposing that the pro- 
cess of evolution of such forms is discontinuous are derived, firstly, 
from the scarcity of observed instances of sudden and large 
variations, while small variations are familiar. Secondly, there 
is a presentiment, which is intuitive in the minds of some, that 
the processes of Nature are continuous processes, and that an 
appearance of discontinuity is due to imperfect knowledge of 
these processes. With the latter difficulty we are not con- 
cerned; but it is in the hope of dispelling the former objection 
that the present observations are recorded. 
It may be remarked that large and sudden variations have not 
unfrequently been observed in organs repeated in a regular 
series, as the petals of regular flowers, &c.; but such changes, 
though considerable, commonly affect all these organs equally 
and in such a way that the original regularity remains in the 
modified structure. The significance of the following examples, 
however, lies in the fact that they not only show the facility 
with which irregular systems may be converted into regular 
ones, but that some of them are also instances of irregular but 
symmetrical systems formed afresh, apparently as sudden varia- 
tions. It will be seen that in some of these cases the resulting 
symmetries, though irregular, are to all appearances as perfect 
as those of the normal flowers. 
Whether the mechanisms of the flowers thus occurring as 
sudden variations are useful mechanisms, and whether they are 
or are not adapted for cross-fertilization by some particular 
insect, we are unable to say. 
