We have met with only one exception, and that only a partial one, namely, with the petioles of the two first leaves of Acanthus candelabrum.  With Delphinium nudicaule the petioles of the two cotyledons are com-[page 554] pletely confluent, and they break through the ground as an arch; afterwards the petioles of the successively formed early leaves are arched, and they are thus enabled to break through the base of the confluent petioles of the cotyledons.  In the case of Megarrhiza, it is the plumule which breaks as an arch through the tube formed by the confluence of the cotyledon-petioles.  With mature plants, the flower-stems and the leaves of some few species, and the rachis of several ferns, as they emerge separately from the ground, are likewise arched.  The fact of so many different organs in plants of many kinds breaking through the ground under the form of an arch, shows that this must be in some manner highly important to them.  According to Haberlandt, the tender growing apex is thus saved from abrasion, and this is probably the true explanation.  But as both legs of the arch grow, their power of breaking through the ground will be much increased as long as the tip remains within the seed-coats and has a point of support.  In the case of monocotyledons the plumule or cotyledon is rarely arched, as far as we have seen; but this is the case with the leaf-like cotyledon of the onion; and the crown of the arch is here strengthened by a special protuberance.  In the Gramineae the summit of the straight, sheath-like cotyledon is developed into a hard sharp crest, which evidently serves for breaking through the earth.  With dicotyledons the arching of the epicotyl or hypocotyl often appears as if it merely resulted from the manner in which the parts are packed within the seed; but it is doubtful whether this is the whole of the truth in any case, and it certainly was not so in several cases, in which the arching was seen to commence after the parts had wholly [page 555] escaped from the seed-coats.  As the arching occurred in whatever position the seeds were placed, it is no doubt due to temporarily increased growth of the nature of epinasty or hyponasty along one side of the part.

As this habit of the hypocotyl to arch itself appears to be universal, it is probably of very ancient origin.  It is therefore not surprising that it should be inherited, at least to some extent, by plants having hypogean cotyledons, in which the hypocotyl is only slightly developed and never protrudes above the ground, and in which the arching is of course now quite useless.  This tendency explains, as we have seen, the curvature of the hypocotyl (and the consequent movement of the radicle) which was first observed by Sachs, and which we have often had to refer to as Sachs’ curvature.