Notes. 
441 
The midrib of the leaf is undoubtedly an organ which partakes of 
the properties of dorsiventral shoots, and the same character conse- 
quently must be inherent to all the fibro-vascular bundles of which it 
is composed, although it remains generally in a latent condition. It 
may, however, for one cause or another, acquire a greater potency and 
lead to a kind of individuation of any one bundle ; which accordingly 
will behave as an independent dorsiventral shoot. This, I think, is 
especially possible in the outermost bundles, which form, on either 
side, the prominent edges of the midrib, on account of their more or 
less isolated and, to a certain degree, independent situation. Now in 
the mango-leaves mentioned, this is precisely what has happened, and 
so their accessory leaves were formed. 
In both cases it is but natural that the epidermis and adjoining 
layers of the leaf-tissues should grow on either side without any 
alteration of structure over the nearest part of the lateral branches 
(or nerves) sprouting forth from the individualized bundle, and there- 
fore the inner surfaces must be the same on either side of the primary 
leaf. 
But then it follows from the polarity of the molecular nature of cells 
and tissues, that the opposite or exterior faces of each accessory leaf 
must be the reverse of the corresponding surfaces of the original 
leaf. 
This explains the somewhat startling fact that a morphologically 
under surface is directed upwards towards the light, and vice versa , 
and gives additional support to the opinion that the difference of 
structure in the two surfaces of the lamina does not depend so much 
on external causes, as on the molecular conditions of the tissues and 
their cellular elements. 
The concave shape of the accessory leaves is the result of unequal 
growth, the lower leaf being hyponastic, the upper one epinastic ; 
because it is evident that the conditions of growth will be more 
favourable for the surface which is the direct continuation of the 
homologous surface of the primary leaf, and this so much the more as 
the geotropism of the dorsal side, and the heliotropism of the ventral 
one, in each case work in the same direction. 
In conclusion, I think we have here a case of fission, but as to its 
primary cause, or causes, I am unable to proffer any suggestion. In 
the first of the two leaves described, there is on the midrib a very 
small knob just where the accessory leaf begins, and it is not 
