From the several cases now given, which refer to widely distinct plants, we may infer that there is some close connection between the reduced size of one or both cotyledons and the formation, by the enlargement of the hypocotyl or of the radicle, of a so-called bulb.  But it may be asked, did the cotyledons first tend to abort, or did a bulb first begin to be formed?  As all dicotyledons naturally produce two well-developed cotyledons, whilst the thickness of the hypocotyl and of the radicle differs much in different plants, it seems probable that these latter organs first became from [page 98] some cause thickened—­in several instances apparently in correlation with the fleshy nature of the mature plant—­so as to contain a store of nutriment sufficient for the seedling, and then that one or both cotyledons, from being superfluous, decreased in size.  It is not surprising that one cotyledon alone should sometimes have been thus affected, for with certain plants, for instance the cabbage, the cotyledons are at first of unequal size, owing apparently to the manner in which they are packed within the seed.  It does not, however, follow from the above connection, that whenever a bulb is formed at an early age, one or both cotyledons will necessarily become superfluous, and consequently more or less rudimentary.  Finally, these cases offer a good illustration of the principle of compensation or balancement of growth, or, as Goethe expresses it, “in order to spend on one side, Nature is forced to economise on the other side.”

Circumnutation and other movements of Hypocotyls and Epicotyls, whilst still arched and buried beneath the ground, and whilst breaking through it.—­According to the position in which a seed may chance to have been buried, the arched hypocotyl or epicotyl will begin to protrude in a horizontal, a more or less inclined, or in a vertical plane.  Except when already standing vertically upwards, both legs of the arch are acted on from the earliest period by apogeotropism.  Consequently they both bend upwards until the arch becomes vertical.  During the whole of this process, even before the arch has broken through the ground, it is continually trying to circumnutate to a slight extent; as it likewise does if it happens at first to stand vertically up,—­all which cases have been observed and described, more or less fully, in the last chapter.  After the arch has grown to some [page 99] height upwards the basal part ceases to circumnutate, whilst the upper part continues to do so.

That an arched hypocotyl or epicotyl, with the two legs fixed in the ground, should be able to circumnutate, seemed to us, until we had read Prof.  Wiesner’s observations, an inexplicable fact.  He has shown* in the case of certain seedlings, whose tips are bent downwards (or which nutate), that whilst the posterior side of the upper or dependent portion grows quickest, the anterior and opposite side of the basal portion of the same internode grows quickest; these two portions