160 REY. PROF. G. HENSLOW ON THE 
as in the Currant. In the Ivy (xxxı., Pl. XXVIII.) the cords of 
all tne whorls occur together, so that the inferior ovary of that 
plant contains on the circumference cords belonging to the calyx, 
corolla, and stamens, as well as the dorsal cords of the carpels, 
the marzinal carpellary cords being near the middle; while all 
that is differentiated of the ovaries is the inner epidermis of each 
ovary-cell with its ovule. 
One can therefore ouly draw a line passing round the dorsal 
cord, including the ovary-cell below it and the two adjacent 
marginal cords as well, to pronounce how much belongs to a 
single carpel. I would not hesitate to call everything included 
within this imaginary circle as foliar ; so that nothing would be 
axial except the boundary, i.e. the epidermis and a certain 
amount of subjacent tissue ; while nothing in the middle what- 
ever would be axial. The above description will be better 
understood by referring to the figures given below; as e. g. 
Calendula (xxxvr., PI. XXIX. p. 184), Gladiolus (Gi, PL 
XXXII, p. 195), &e. 
THE COHESION or CARPELS. 
The important feature of the arrest of development which takes 
place whenever two organs are congenitally united, cannot be too 
emphatically emphasized, as much confusion has arisen in con- 
sequence of a want of recognition of this fact, especially with 
regard to the nature of the receptacular tube. This arrest will 
also quite interpret the origin of the basilar ovule as well as of 
the free-central placenta of the Primulacez *. 
Parietal Placentations.—As might be anticipated, there is 
differenee in the distribution of the cords of carpels which are 
provided with parietal and axile placentas respectively. In the 
latter case the rule is for the outermost branch of the triple group, 
or of the “ horseshoe ”-shaped group belonging to each carpel, 
to become tke dorsal cord, and the ovary-cell to appear between 
it and the other two remaiuing portions, which may be fused to- 
gether to form one of the central axial or marginal carpellary 
cords (e. g, Gladiolus, xx. ; Pl. XXXII.). 
When, however, the carpels have a parietal placentation, as in 
Reseda fr PI. XXIV.), after supplying the stamens, å more or 
* As these various results of arrest of development under cohesion have 
been fully discussed, I must refer the reader to my book, ‘ Floral Structures,' 
for details: see Index, “ Arrest.” 
