In some other cases the hypocotyl or epicotyl protrudes from the seed at first only slightly bowed; but the bowing afterwards increases independently of any constraint. The arch is thus made narrow, with the two legs, which are sometimes much elongated, parallel and close together, and thus it becomes well fitted for breaking through the ground. With many kinds of plants, the radicle, whilst still enclosed within the seed and likewise after its first protrusion, lies in a straight line with the future hypocotyl and with the longitudinal axis of the cotyledons. This is the case with Cucurbita ovifera: nevertheless, in whatever position the seeds were buried, the hypocotyl always came up arched in one particular direction. Seeds were planted in friable peat at a depth of about an inch in a vertical position, with the end from which the radicle protrudes downwards. Therefore all the parts occupied the same relative positions which they would ultimately hold after the seedlings had risen clear above the surface. Notwithstanding this fact, the hypocotyl arched itself; and as the arch grew upwards through the peat, the buried seeds were turned either upside down, or were laid horizontally, being afterwards dragged above the ground. Ultimately the hypocotyl straightened itself in the usual manner; and now after all these movements the several parts occupied the same position relatively to one another and to the centre of the earth, which they had done when the seeds were first buried. But it may be argued in this and other such cases that, as the hypocotyl grows up through the soil, the seed will almost certainly be tilted to one side; and then from the resistance which it must offer during its further elevation, the upper part of the hypocotyl will be doubled down and thus become arched. This view seems the more probable, because with Ranunculus ficaria only the petioles of the leaves which forced a passage through the earth were arched; and not those which arose from the summits of the bulbs above the ground. Nevertheless, this explanation does not apply to the Cucurbita, for when germinating seeds were suspended in damp air in various positions by pins passing through the cotyledons, fixed to the inside of the lids of jars, in which case the hypocotyls were not subjected to any friction or constraint, yet the upper part became spontaneously arched. This fact, moreover, proves that it is not the weight of the cotyledons which causes the arching. Seeds of Helianthus annuus and of two species of Ipomœa (those of ‘I. bona nox’ being for the genus large and heavy) were pinned in the same manner, and the hypocotyls became spontaneously arched; the radicles, which had been vertically dependent, assumed in consequence a horizontal position. In the case of Ipomœa leptophylla it is the petioles of the cotyledons which become arched whilst rising through the ground; and this occurred spontaneously when the seeds were fixed to the lids of jars.