is not arched, as it has not to force for itself a passage through the ground.  In the accompanying sketch (Fig. 58) the petiole of the first leaf has already partially straightened itself, and the blade is beginning to unfold.  The small second leaf ultimately grows to an equal size with the first, but this process is effected at very different rates in different individuals:  in one instance the second leaf did not appear fully above the ground until six weeks after the first leaf.  As the leaves in the whole family of the Acanthaceae stand either opposite one another or in whorls, and as these are of equal size, the great inequality between the first two leaves is a singular fact.  We can see how this inequality of development and the arching of the petiole could have been gradually acquired, if they were beneficial to the seedlings by favouring their emergence; for with A. candelabrum, spinosus, and latifolius there was a great variability in the inequality between the two first leaves and in the arching of their petioles.  In one seedling of A. candelabrum the first leaf was arched and nine times as long as the second, which latter consisted of a mere little, yellowish-white, straight, hairy style.  In other seedlings the difference in length between the two leaves was as 3 to 2, or as 4 to 3, or as only .76 to .62 inch.  In these latter cases the first and taller leaf was not properly arched.  Lastly, in another seedling there was not the least difference in size between the two first leaves, and both of them had their petioles straight; their laminae were enfolded and pressed against each other, forming a lance or wedge, by which means they had broken through the ground.  Therefore in different individuals of this same species of Acanthus the first pair of leaves breaks through the ground by two widely different methods; and if [page 80] either had proved decidedly advantageous or disadvantageous, one of them no doubt would soon have prevailed.

Asa Gray has described* the peculiar manner of germination of three widely different plants, in which the hypocotyl is hardly at all developed.  These were therefore observed by us in relation to our present subject.

Delphinium nudicaule.—­The elongated petioles of the two cotyledons are confluent (as are sometimes their blades at the base), and they break through the ground as an arch.  They thus resemble in a most deceptive manner a hypocotyl.  At first they are solid, but after a time become tubular; and the basal part beneath the ground is enlarged into a hollow chamber, within which the young leaves are developed without any prominent plumule.  Externally root-hairs are formed on the confluent petioles, either a little above, or on a level with, the plumule.  The first leaf at an early period of its growth and whilst within the chamber is quite straight, but the petiole soon becomes arched; and the swelling of this part (and probably of the blade) splits open one side of the