[Footnote 36: Sophian, quoted by Bochart in his Hierozoïcon, ii. cap. xxi, p. 497.] The argument may be stated thus: seeds remain for months undecayed, and still capable of germination, at depths varying from one to twenty inches below the surface of the soil in certain ants' nests, why should they not lie hidden for indefinite periods in ordinary soil? To answer this positively, experiments should be made[37] in order that we might learn whether these seeds can retain their vitality without sprouting in moist soil; but the general belief is that under these conditions they will do one of two things, they will either grow or rot. Be this as it may, one of the most curious points that we have learned about these ants, is that they know how to preserve seeds intact, even when within from one to three inches of the surface of the ground, that is to say, at the actual depth at which a gardener most frequently sows his seeds, though if these very seeds are taken out of the granary and sowed by hand, they will germinate in the ordinary way. It is possible that this may be in part due to the compact nature of the floors and ceilings of the granaries, these excluding air in some measure, though as moisture freely passes through them, and there are always two or three open galleries leading into the granaries, and which communicate directly with the open air, I can scarcely accept this explanation as complete. [Footnote 37: In order to try the experiment fairly, seeds taken from ants' nests, or seeds of the same species as those which are habitually found in ants' nests, should be placed at different depths in the earth and examined after the lapse of six or eight months. Why it is that certain seeds resist the influences which destroy the vitality of other seeds of closely allied species is another and a very curious but complicated problem, the explanation of which may perhaps lie in the different chemical properties of the seeds in question, in the more or less permeable character of their seed-coats, or their general texture.] The seeds do occasionally sprout in the nest, though it is extremely rare to find instances of this, and then the ants nip off the little root, and carry each seed out into the air and sun, exactly as the old writers have described, and when the growth has been checked and the seed malted by exposure, they fetch them in again. It is in this condition that the ants like best to eat them, as I have proved by experiments among my captives.