62.2 PLANTS JAVANIC^} RARIORES.
regard to the state of the unimpregnated ovulum, which I have some reason to believe is not orthotropous as might be expected, and as it has been described, but apparently anatropous, and that perhaps in the whole tribe. As, however, my observations on this subject are entirely made from the macerated ovaria of dried specimens, the state- ment here made must be received as requiring confirmation from the examination of living plants, aiid of a greater number of species. 1
From this ordinary direction of embryo in the tribe, the deviations are of two kinds : the first, and no doubt the more important, is that in which the radicle is placed at a point close to the umbilicus, which is the most general structure in Phoenogamous plants ; but as it never points directly within the umbilicus, either in this or any other family, I have modified the expression generally employed in such cases. The second deviation is where the umbilicus is placed on or near the middle of the ripe seed with the radicle pointing to its lower extremity; in other words, where the embryo is parallel to the umbilicus. But this position of umbilicus of the ripe seed does not necessarily imply an exactly similar insertion in the unimpregnated ovulum ; and in this tribe I am inclined to believe that in many cases the foramen of the ovulum is so close to the 225] umbilicus as to appear anatropous, and that it ulti- mately becomes more distant from the unequal growth of the opposite extremities of the seed.
The exceptions to the ordinary structure in Sterculiece which appear to be next in importance are the modifications in texture, and especially in the period of dehiscence of the seed-vessel or carpel, or even its nondehiscence, for in this respect the tribe admits of the two extremes. In the first, where the carpel opens long before the ripening of the seed, its texture is always foliaceous, and the embryo may
1 The species of Stercidia with orthotropous embryo in which I have found this unexpected position of foramen in the unimpregnated ovulum, avefcetida, guttata, cart hag ineusis, nobilis, and angustifolia ; and in the ripe seeds of traga- canthce, wens, villosa, and quadrifida, an indication of a lateral foramen near the base is still visible, but which mfcetida I have not been able to detect
�� �
