62 
W. C. Worsdell. 
forms shews, results from the complete branching into two, of one 
of the members of the whorl. In the Wallflower it is not infre¬ 
quent for the ovary to become quadri-instead of bi-carpellary, 
which is probably a reversion to the primitive condition in which 
the carpels were equal in number to the petals, sepals and inner 
stamens. Many Dicotyledonous flowers whose parts are in fours, 
are apt to become pentamerous, and in many pentamerous flowers I 
occasionally see an extra member or two added to each whorl. 
In all flowers we may regard negative dedoublement or reduc¬ 
tion of parts as a progressive , and positive doubling or an increase 
of parts as a retrogressive or reyersional phenomenon. 
To cite one case in which positive dedoublement takes place in 
foliage leaves, Lonicera Periclymenum normally produces two leaves 
in each whorl; occasionally, however, shoots occur bearing whorls 
of four leaves; the normal habit is probably the primitive one; 
similar cases occur in many other plants. 
Twins, whether vegetable or animal, doubtless belong to this 
category. The fertilized egg-cell, instead of initiating its develop¬ 
ment into an embryo by means of a transverse division of its 
contents, which is the normal case, might divide by a vertical 
wall. Each sister-cell thus formed might then develope into an in¬ 
dependent embryo. This would be congenital positive dedoublement, 
for the production of a single embryo is the normal and primitive 
character. In Loranthus sphaerocarpus , described by Treub, the 
fertilized ovum, divides by a vertical wall but the sister-cells 
develope together into a single pro-embryo, consisting of a double 
row of cells. Twin-flowers, if we regard the individual flowers them¬ 
selves and not the shoot as a whole which bears them, afford 
another case in point. 
III.— Nkutral Condition. 
All the phenomena cited under this head may be described as 
cases of post-genital positive doubling; this refers, of course, merely 
to their ontogenetic history. From the point of view of their 
phylogeny, however, much more is involved, and hence I place 
them under a distinct heading for separate discussion. The 
character, described in its simplest terms, which is common to all 
these cases, is this: that the organ or tissue at its first origin an 
integrity, becomes later branched or sub-divided in its iipper or 
younger region. But I will postpone the consideration of the 
