32 
undisputed affinity with bulbs, whether tunicate or Squamate, 
it is impossible not to admit the existence at the place in ques- 
tion of several very short internodia. Agardh himself, con- 
sidering the fruit as a terminal bud, consisting in its simplest 
form of one carpellary leaf with the seed-bearing stalk in its 
axilla, does not by this view embrace the greater portion of 
the phenomena. According to my observations in many 
compound fruits, whether unilocular or plurilocular, with a 
so called central or columnar placenta, there does not arise a 
separate bud from the axilla of each carpellary leaf, but the 
whole central support or common seed-stalk is the immediate 
prolongation of the floral axis or peduncle, forming an inter- 
nodium above the nodus from whence proceed the carpellary 
leaves. It is then either covered with seeds dispersed with- 
out any perceptible order, as in Primulacee and Caryophyl- 
lacew, or it forms at a greater or less height a special nodus 
from whence one or two seeds descend into each cell of the 
fruit, supported on distinct seminal pedicels, as in the Malva- 
cee with monospermous carpels, Phytolacca decandra, and all 
Euphorbiacexw. In the latter family the seed-stem is already 
almost free, that is, the margins of the carpellary leaves 
scarcely adhere to it, and readily separate when ripe without . 
any laceration of the tissue. 
In the Boraginez the seed-stem as is known is entirely 
free, or on the outside of four completely closed carpels ar- 
ranged around on one horizontal plane. 
fi hese closed carpels or caryopses have in their structure, 
especially as regards their outer integuments, considerable 
resemblance to those seeds which Mirbel* distinguishes 
under the name of curved cr campylotropous, inasmuch as 
the aperture for the passage of the distinct portion in seminal 
pedicels (Spermopodium mihi non Hoffm. Funiculus umbili- 
calis auct.) is very near to the organic base of the carpellary 
leaf, wherefore they are generally spoken of by descriptive 
Botanists as Nuces basi perforate, and wherefore the error of 
Linneus, who considered ‘them as naked seeds, was excus- 
able. Thus, in the genus Asperugo, each carpel has the 
appearance of a free pendulous coccus, which the analogy 
_ _* Nouvelles recherches sur le structure de l’ovule végétale et sur ses 
développemens. 1. Mem. 1828 p- 3 
