Animal Intelligence / The International Scientific Series, Vol. XLIV.

by George John Romanes

From the disk in various directions there radiate ant-roads or avenues, which are cleared and smoothed like the disk itself, and which course through the thick surrounding grass, branching and narrowing as they go till they eventually taper away. These roads are usually three or four in number before they begin to branch, but may be as many as seven. They are usually two to three inches wide at their origin, but in large nests may be as much as five. MacCook found no road longer than sixty feet, but Lincecum describes one of three hundred feet. Along these hard and level roads there is always passing, during the daytime of the harvesting season, a constant stream of ants—those going from the nest being empty-handed, and those returning to it being laden with seeds. Of course the incoming ants, converging from all quarters upon the road, and therefore increasing in numbers as they approach the nest, require greater space for free locomotion; while the outgoing ants, diverging as they get further from home, also require greater proportional space the less their distance from the nest: hence the gradual swelling in the width of the roads as they approach the nests.

The manner of collecting the seeds in the jungle surrounding the roads is thus described by MacCook:—

At last a satisfactory seed is found. It is simply lifted from the ground, or, as often happens, has to be pulled out of the soil into which it has been tightly pressed by the rain or by passing feet. Now follows a movement which at first I thought to be a testing of the seed, and which, indeed, may be partially that; but finally I concluded that it was the adjusting of the burden for safe and convenient carriage. The ant pulls at the seed-husk with its mandibles, turning and pinching or 'feeling' it on all sides. If this does not satisfy, and commonly it does not, the body is raised by stiffening out the legs, the abdomen is curved underneath, and the apex applied to the seed. I suppose this to be simply a mechanical action for the better adjusting of the load. Now the worker starts homeward. It has not lost itself in the mazes of the grass forest. It turns directly towards the road with an unerring judgment. There are many obstacles to overcome. Pebbles, pellets of earth, bits of wood, obtruding rootlets, or bent-down spears of grass block up or hinder the way. These were scarcely noticed when the ant was empty-handed. But they are troublesome barriers now that she is burdened with a seed quite as thick, twice as wide, and half as long as herself. It is most interesting to see the skill, strength, and rapidity with which the little harvester swings her treasure over or around, or pushes it beneath these obstacles. Now the seed has caught against the herbage as the porter dodges under a too narrow opening. She backs out and tries another passage. Now the sharp points of the husk are entangled in the grass. She jerks or pulls the burden loose, and hurries on. The road is reached, and progress is comparatively easy. Holding the grain in her mandibles well above the surface, she breaks into what I may describe with sufficient accuracy as 'a trot,' and with little further interruption reaches the disk and disappears within the gate. There are variations from this behaviour, more or less marked, according to the nature of the grounds, the seeds, and (I suppose) the individuality of the harvesters; but the mode of ingathering the crop is substantially as above. Each ant operated independently. Once only did I see anything like an effort to extend sympathy and aid. A worker minor seeming to have difficulty in testing or adjusting a large seed of buffalo-grass, was assisted (apparently) by one worker major, and then by another, after which she went on her way.

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.

The other main point on which his observations are defective has reference to a remarkable statement made by Lincecum in the most emphatic terms. This statement is that upon the surface of their disk the ants sow the seeds of a certain plant, called ant-rice, for the purpose of subsequently reaping a harvest of the grain. There is no doubt that the ant-disks do very often support this peculiar kind of grass, and that the ants are particularly fond of its seed; but whether the plant is actually sown in these situations by the insects, or grows there on account of these situations being more open than the general surface of the ground—this question MacCook has failed to answer, or even to further. We are, therefore, still left with Dr. Lincecum's emphatic assurance that he has witnessed the fact. His account is that the seed of the ant-rice, which is a biennial plant, is sown in time for the autumnal rains to bring up. At the beginning of November a green row or ring of ant-rice, about four inches wide, is seen springing up round the circumference of the disk. In the vicinity of this circular ring the ants do not permit a single spire of any other grass or weed to remain a day, but leave the aristida, or ant-rice, untouched until it ripens, which occurs in June of the next year. After the maturing and harvesting of the seed, the dry stubble is cut away and removed from the pavement or disk, which is thus left unencumbered until the ensuing autumn, when the same species of grass again appears as before, and so on. Lincecum says he has seen the process go on year after year on the same ant-farms, and adds,—

There can be no doubt of the fact that the particular species of grain-bearing grass mentioned above is intentionally planted. In farmer-like manner the ground upon which it stands is carefully divested of all other grasses and weeds during the time it is growing. When it is ripe the grain is taken care of, the dry stubble cut away and carried off, the paved area being left unencumbered until the ensuing autumn, when the same 'ant-rice' reappears within the same circle, and receives the same agricultural attention as was bestowed upon the previous crop—and so on year after year, as I know to be the case, in all situations when the ant's settlements are protected from graminivorous animals.