VASCULAR SYSTEMS OF FLORAL ORGANS. 161 
less complete ring is formed by “ closing up.” This, then, breaks 
up into two, three, or more groups for the placentas, with interme- 
diate smaller cords for the dorsal. In Ribes (xxv. 5; Pl. XX VILL), 
which has a receptacular tube, there are four additional cords 
present besides the dorsal and marginal carpellary. These four 
supply part of the perianth and are inserted in alternating posi- 
tions with the four carpellary cords. This may be likened to the 
way in which a leaf of this plant is inserted on the branch. It 
has three cords in the petiole which pass down the axial cylinder, 
while intermediate ones pass up the stem between them. 
Certain anomalies have long been noticed in the structure of 
the fruit of Papaveracee and Crucifere, especially as to the 
character of the placentas with the stigmas superposed to them. 
To endeavour to arrive at some interpretation of the anomalies, 
we must compare them with the behaviour of an original or po- 
tential carpellary cord while within an axis. For at this point 
we find marked differences between the cords of these orders and 
those of others, such, for example, as of Hellebore, Eranthis, or 
Aconite. In these latter a cord in the axis, which is destined to 
give rise to a carpel, divides into three branches, one being dorsal, 
the other two marginal or placentary, while the ovary-cell appears 
between them (Eranthis, tv. 6-9; Pl. XXIIL). In Papaver (v.) 
and Eschscholtzia (vx. ; Pl. XXIII.) a certain number of cords 
appear which at once supply the placentas and do nothing else. 
They do not divide into triplets ; nor, if we compare these with 
Viola (vırı.) and Reseda (1x.; Pl. XXIV.), which have parietal 
placentas, do we find, as in them, distinct dorsal cords. In these 
latter genera, besides the presence of large placentary cords, there 
are smaller dorsal ones alternate in position with them. Hence it 
would seem that in Cruciferous and Papaveraceous plants we 
have an extremely arrested or undifferentiated condition of the 
carpellary cords; so that the large cords of the Wallflower which 
stand between the longer pair of stamens (vir. 7) only develop 
into a pair of closely approximated placental cords with no dif- | 
ferentiated dorsal cord at a!l. Hence the carpellary leaf which 
this cord ought to belong to would, theoretically, assume the 
form of the needle-like leaf of a Pine; but only possessing two | 
marginal cords and no dorsal one (like the petal of a Composite). 
The consequence is that the stigma arising from these two 
approximate foliar edges will necessarily stand directly over the 
placentas. 
