GENERAL REMARKS. 131 
upper ones pass insensibly almost into anthers, these are 
plurilocular and the cells are either approximated in pairs or 
single. 
In Calla aromatica, the organisation is a good deal different. 
In the first place we have the spatha of Pothos, for there is 
no greater degree of involution or of development at its 
base than else where. 
We have the tendency to tricarpellarity, and what is very re- 
markable in the order is the suffultion of each ovary by a 
bract, which takes on the appearance of a gland, or rather 
abortive anther, but is very regular in its situation. 
The anthers again approximate much to those of Pothine, 
having a distinct flattened membranous brown filament, 
which adheres to the spadix, and a bilocular typical extrorse 
anther, they are almost always aggregated in threes, the third 
being in such cases the lower. 
In the venation of leaves it approaches Pothos, which is 
much more Monocotyledonous than Caladium or Arum. Calla 
is itself truely Monocotyledonous in this respect. 
Another singular thing is the fasciculate, or aggregate in- 
florescence in one axilla, and the decidedly terebinthenous 
odour, from which I would infer the presence of Coniferous, 
or rather resiniferous, glands or dots. 
The ovula in their number and form are much more ty- 
pical, and approach again to Pothos, the foramen being near 
the hilum. 
Does it not present another affinity to Pothos, by the sheaths 
of the peduncles of the spathes. Philodendron passes into 
Pothos directly by its habit. 
In Caladium Colocasia, each young leaf is enveloped іп а 
simple sheath, which is not, besides, strictly axillary. The ve- 
nation is conducted on thesame principles, but the outer lamina 
after having made the complete circuit, is inflected for a short 
distance between the midrib and itself; see Fig. Па,. Pl. 
CLXXIII. This inflexion however only takes place near 
base of the leaf. 
