496 ON AN IMPROVED CLASSIFICATION OF FRUITS. 
along its border. Although the production of germs from the axis, 
as maintained by some high botanical authorities, is not antecedently 
very improbable, I cannot consider it as established by any good 
evidence, and it supposes so remarkable an anomaly in the mode of 
fertilization, as cannot be admitted without certain proof. Again, 
each carpel, according to the analogy of the leaf, has an upper and 
under surface, and a middle portion containing the vascular system. 
The under surface, which forms the external covering of the fruit, is 
called the epicarp ; the vascular layer, the mesocarp ; and the upper 
surface, which lines the interior of the carpel, the endocarp. The 
differences in the mode of development of these parts, explain the 
membranous, coriaceous, woody, fleshy, or pulpy character of fruits, 
or certain portions of them; and it thus appears why these differ- 
ences are of minor importance, and may occur between fruits of the 
same essential structure. 
The number of germs produced in a carpel depends partly on the 
productive tendency inherent in the species, arising from its elemental 
structure and vital energy ; much, also, on the space afforded to it 
and the amount of nutriment it receives. It is common for a carpel 
to be single-seeded, and not uncommon for the seed so closely to fill the 
folded carpel that the whole passes for a naked seed. It may have two 
or several seeds; and in a few instances the germ-producing or pla- 
cental portion of the edge of the carpel is extended and crowded with 
germs so as greatly to multiply the seeds. The coherence of carpels 
in a circle is very common, and may either be slight and partial, pro- 
ducing a lobed fruit, or more complete—either by the edges only of 
the carpels, causing a one-celled capsule with parietal placente, by 
their meeting on the axis so.as to cause axillary placente, or by 
their turning inward from the axis, so that the placentz project into 
the cells; and in these cases, if the substance be membranous, coria- 
ceous, or woody, the opening may be by the separation of the carpels, 
by the splitting asunder of the midrib of the carpel, by separation of 
the external portion from the firmly united imfolded parts, by the 
turning back of valves at the upper part, by circumscission, or by pores 
formed to allow of the escape of the seeds. If we add to these cir- 
cumstances the various adherences of exterior parts, we have the 
_ means of explaining the nature of all known fruits. We endeavour 
to express the facts with as many distinct names for varieties of fruits. 
as we have found adopted by good authorities, and can perceive to be 
