But these ants do not confine their harvesting operations to gathering fallen seeds; they will, like the ants of Europe, also cut seeds from the stalk. In order to test the disposition of _crudelis_ to garner the seeds from the stem, bunches of millet were obtained from the North, and stalks eighteen inches high, crowned by the boll of close-set seeds, were stuck in the mound of an active formicary. The ants mounted the stems and set to work vigorously to secure the seeds, clusters of twenty or more being engaged at once upon one head. The seeds were carried off and stored within the nest. This experiment proved pretty conclusively that in the seeding season _crudelis_ does not wait for the seeds to drop, but harvests them from the plant. The 'granaries' into which the seeds are brought are kept distinct from the 'nurseries' for the pupæ. Their walls, floor, and roof are so hard and smooth, that MacCook thinks the insects must practise upon them 'some rude mason's craft.' He traced these granaries to a depth of four feet below the surface of the ground, and believes, from the statements of a native peasant, that they, or at least the formicaries, extend to a depth of fifteen feet. As regards the care that the ants take of the gathered grain, Lincecum describes the same habit as Moggridge and Sykes describe--viz., the sunning of wet seeds to dry. MacCook, however, neglected to make any experiments on this subject. Neither has he been able to throw any light upon the question as to why the stored seeds do not germinate, and is doubtful whether the habit of gnawing the radicle of sprouting seeds, which prevails in the European species, is likewise practised by the American. On two other points of importance MacCook's observations are also incomplete. One of these has reference to an alleged statement, which he is disposed to believe, that when some ants in a community have been killed by poison, the survivors avoid the poison: he, however, made no experiments to test this statement.