y in many species of _Vaccinium_.

A division of the individual carpels occurs very frequently when those
organs become more or less leafy, as in _Trifolium repens_, and other
plants to be hereafter mentioned.

The instances given in this chapter have all been cases wherein the
division or the accessory growth has taken place in one plane only and
that plane the same as that of the affected organ, but there are other
examples, probably equally due to fissiparous division, where the new
growth is either parallel to, or even at angle with the primary organ.
Of such nature are some of those instances wherein two leaves appear to
be placed back to back. These partake of the nature of excrescences or
of exaggerated developments, and hence will be more fully treated of
under the head of hypertrophy. It must be remembered that in some of
these cases the fission may be a resumption of characters proper to the
species under natural conditions, but lost by cultivation or otherwise.
Thus, Mr. Buckman accounts for "finger-and-toe" in root-crops on the
principle of reversion to the wild form.

FOOTNOTES:

[67] Loc. cit., p. 295.

[68] 'Bull. Soc. Bot. France,' 1857, p. 758.

[69] Masters, 'Jourl. Linn. Soc.,' vol. vii, p. 121.

[70] Cramer, 'Bildungsabweichungen,' p. 4, tab. vi, fig. 4, figures a
case of the same kind in _Pinus Cembra_.

[71] 'Ann. des Science Nat.,' 2nd series, t. iv, p. 147, tab. v, figs. 3
and 4.

[72] 'Mem. Acad. Scien. Toulouse,' 5th series, vol. iii.

[73] Duchartre, 'Ann. Sc. Nat.,' 3rd series, 1848, vol. x, p. 207.

[74] Masters, 'Rep. Bot. Congress,' London, 1866, p. 136, tab. 7, f. 15,
16.

CHAPTER II.

DIALYSIS.

This term is here made use of in the same sense as in descriptive
botany, to indicate the isolation of parts of the same whorl; it is thus
the opposite of cohesion. Morren, as has been previously stated,
employed the word in a different sense, while Moquin-Tandon[75] included
cases of this description under the category of "Disjonctions qui
isolent les organes."

Dialysis, as here understood, may be the result of an arrest of
development, in consequence of which parts that under ordinary
circumstances would become fused, do not do so; or, on the other hand,
it may be the result of an actual separation between parts primitively
undivided. As it is not possible in every case to distinguish between
the effects of these two diverse causes, no attempt is here made to do
so