The gardens of Apterostigma are provided with a special mycelial envelope, but they are naked in all the other Attiine genera and subgenera. Mællerius and Acromyrmex make a single large garden on the floor of the single nest-chamber. Finally, in Atta s. str., the nest attains huge dimensions and comprises a number of large chambers, each with its sessile fungus-garden resembling the single garden of Acromyrmex and Mallerius.

While this series of forms shows an interesting and significant advance in the methods of raising fungi, it throws little light on the origin of this complicated and extraordinary habit. Forel is inclined to believe that the ancestral Attii lived, like the existing Apterostigma, in rotten wood and gradually acquired the habit of cultivating on insect excrement the fungi which they chanced to find in the moist galleries of their nests. Von Ihering surmises that the Attii are descended from harvesting ants which transferred their appetite from the hard seeds in their chambers to the delicate fungi that accidentally grew on these In this connection Santschi's observations on the seed-storing Oxyopomyrmer, cited on p. 273, are very suggestive. Besides the Attii there are also two other groups of fungus-growing insects, the ambrosia beetles (Scolytidae), which cultivate fungus on their excrement or on the walls of their burrows in the trunks of trees, and a long series of paleotropical termites which raise fungi on sponge-like masses of their excrement in large chambers like those of Atta s. str. A study of these cases and of the female Atta while she is establishing her colony, would seem to indicate that all fungusgrowing insects originally used their own excrement as a substratum. for their gardens and only later took to adding other substances (excrement of other insects and pieces of leaves in the Attii, wood-shavings in the ambrosia beetles). But how these various insects first came into possession of the fungi which they now so assiduously cultivate and transmit from generation to generation, we are unable to state, especially as our knowledge of these plants is still so rudimentary that we cannot even say whether they are Ascomycetes or Basidiomycetes, independent species or merely permanently or temporarily modified phases of certain well-known moulds or mushrooms. The study of the Attii and other fungus-growing insects has only just begun, and further advance in this fascinating subject will be more difficult for the mycologist than for the entomologist. The latter, however, will have to build. on the investigations of the former.

CHAPTER XIX.

THE RELATIONS OF ANTS TO PLANT-LICE, SCALE-INSECTS, TREE-HOPPERS AND CATERPILLARS.

"Hæ formicarum vaccæ."-Linné, "Systema Naturæ," Ed. 12, I, 1766.

"Viennent enfin ces peuplades qui couvrent la surface de la terre, et dont les républiques sont si nombreuses que le globe ne leur suffiroit pas si la nature n'eût mis de justes bornes à leur multiplication. Une foule d'insectes deviennent leur proie; la petitesse des individus est compensée chez elles par le nombre, mais la force n'est pas leur principale ressource. Ce ne sont pas non plus les fleurs et les fruits qui leur fournissent leur pâture ordinaire; elle est l'objet d'une industrie plus recherchée. Les peuplades dont nous parlons, vont la recueillir auprès de certains êtres pacifiques qui vivent en troupes, et leur prodiguent sans contrainte les sucs qu'ils savent extraire des plantes. Elles ont l'art de s'en faire entendre, de les réunir dans leur habitation, et de les défendre contre leurs ennemis."-P. Huber, "Recherches sur les Mœurs des Fourmis Indigènes," 1810.

In gaining their wide and intimate acquaintance with the vegetable world the ants have also become acquainted with a large number of insects that obtain their nutriment directly from plants, either by sucking up their juices or by feeding on their foliage. To the former group belong the phytophthorous Homoptera, the plant-lice (Aphididæ),. scale insects or mealy bugs (Coccida), tree-hoppers (Membracidae), lantern-flies (Fulgoridæ) and jumping plant-lice (Psyllida); to the latter belong the caterpillars of the Lycaenid butterflies, the "blues," or "azures" as they are popularly called. All of these creatures excrete liquids which are eagerly sought by the ants and constitute the whole, or at any rate, an important part of the food of certain species. In return the Homoptera and caterpillars receive certain services from the ants, so that the relations thus established between these widely different insects may be regarded as a kind of symbiosis. These relations are most apparent in the case of the aphids, and as these insects have been more often and more closely studied in Europe and America, they may be considered first and at somewhat greater length.

The consociation of the ants with the aphids is greatly facilitated by the gregarious and rather sedentary habits of the latter, especially in their younger, wingless stages, for the ants are thus enabled to obtain a large amount of food without losing time and energy in ranging far afield from their nests. Then, too, the ants may establish their

nests in the immediate vicinity of the aphid droves or actually keep them in their nests or in "sheds" carefully constructed for the purpose.

Nearly all plants, except the cryptogams, may be infested by aphids, and no part of the plant is free from their attacks. Certain species prefer the leaves, others the twigs, and still others the roots and subterranean stems. Most species live on the surfaces of the plants but a number also make and inhabit galls. Only the former, of course, are accessible to the ants. The sedentary and gregarious habits of the aphids also expose them to a host of enemies, among which the Coccinellid beetles and their larvæ, the larvæ of certain Diptera (Syrphidae) and Neuroptera (Chrysopa, Hemerobius) and a host of small parasitic Hymenoptera (Pteromalidæ, Braconidæ, Crabronida) are the most formidable.

The aphids pierce the integument of the plant with their slender, pointed mouth-parts and imbibe the juices which consist of water con

FIG. 204. Portion of tap-root of wormwood (Artemisia vulgaris) with Lasins umbratus workers attending young and adult plant-lice (Trama radicis). (Mordwilko.) a, Sucking adult Trama with hind legs raised; b, ant palpating Trama with her feelers; c, ant receiving droplet of honey-dew from anus of Trama; d and e, ants carrying away plant-lice in their jaws; f, Trama with droplet of honey-dew suspended from perianal hairs.

taining in solution cane sugar, invert sugar, dextrin and a small amount of albuminous substance. In the alimentary tract of the insects much of the cane sugar is split up to form invert sugar, and a relatively small amount of all the substances is assimilated, so that the excrement is not only abundant but contains more invert and less cane sugar than the juices of the plant. This excrement is voided in colorless drops, and when it falls on the leaves of the plants and dries in the air is known as "honey-dew," the ros melleus, mel aërium, roscida mella, melligo, pét of the ancient writers. Réaumur (1737) and Leche (1765) seem to have been the first to ascertain that the honey-dew, which the ancients supposed to come from the plants, from the sky or

from some other mysterious source, is voided by the droves of plantlice. Within more recent years this subject has been exhaustively studied by Büsgen (1891). The quantity of honey-dew excreted by the aphids, when we consider the small size of these insects, is most surprising. Büsgen found that a single linden aphis excretes nineteen drops in twenty-four hours, while the maple aphis excretes as many as forty-eight drops during the same

period. A source of nutriment at once so rich and so inexhaustible, could hardly remain unnoticed and unexploited by the ants in their interminable search for liquid food.

Some ants (Leptothorax sp.) obtain the honey-dew merely by licking the surface of the leaves and stems on which it has fallen, but many species have learned to stroke the aphids and induce them to void the liquid gradually so that it can be imbibed directly. A drove of plant-lice, especially when it is stationed on young and succulent leaves or twigs, may produce enough honey-dew to feed a whole colony of ants for a considerable period. But the aphidicolous habit has not been acquired by all ants. The highly carnivorous Doryline and Ponerinæ never attend these insects and care nothing for their excretions, and the same is true of the exquisitely carnivorous, granivorous and fungus-eating genera of Myrmicinæ (Pseudomyrma, Pogonomyrmex, Atta, etc.). Other Myrmicine genera, such as Myrmica, Cremastogaster, Tetramorium and Monomorium contain many aphidicolous species. The two highest subfamilies, however, the Dolichoderinæ and Camponotinæ, represent the most perfect development of the habit, especially the genera Iridomyrmex, Dolichoderus, Azteca and Liometopum of the former, and Lasius, Brachymyrmex, Prenolepis, Plagiolepis, Ecophylla, Formica, Myrmecocystus and Lasius of the latter subfamily. In our north temperate region the species of Lasius

FIG. 205.

excell those of all other genera in this respect; in fact, the yellow, subterranean ants of the subgenus Acanthomyops, which is peculiar to North America, live exclusively on the excrement of root-aphids and coccids.

The behavior of ants in the presence of aphids has been observed by P. Huber (1810), Forel (1874), Witlaczil (1882), Lubbock (1888a), Büsgen (1891), Lichtenstein (1877-'80), Forbes (1894. 1905, 1906), Del Guercio, Kolbe (1888), Shouteden (1902), Mordwilko (1896, 1901, 1907), and others too numerous to cite. One of the best of the early accounts of this behavior is that of Huber (p. 181 et seq.): "A thistle branch was covered with brown ants [Lasius niger] and aphids. I observed the latter for some time, in order, if possible, to ascertain the precise moment when they emitted this secretion from their bodies; but I remarked that it exuded very rarely of itself, and that the aphids, when separated from the ants, discharged it to a distance, by making a movement like a sudden jerk.

"Why did nearly all the ants that were climbing about on the stems, have their abdomens distended as if with some liquid? This question I was able to answer by watching a single ant, whose exact method of procedure I will endeavor to describe. I saw her first crawl over some aphids without pausing and without disturbing them; but she soon halted near one of the smallest and seemed to caress its abdomen, stroking it alternately first with one and then with the other antenna. I was surprised to see the liquid escape from the aphid's body, and the ant seize and imbibe the droplet at once. Her antennæ were thereupon applied to another much larger aphid, which, on being caressed in the same manner, voided a larger drop of the nutrient liquid. The ant advanced to seize it, and then moved on to a third which she caressed in the same manner. The liquid was voided immediately and received by the ant. She moved on; a fourth, probably already exhausted, refused to respond to her solicitations and the ant, probably divining that she had nothing to expect, quitted this aphid for a fifth from which I saw her obtain a further supply of food.

"A few such repasts are quite sufficient, and the satiated ant returns to the nest. Thereupon I watched the other ants that had remained. behind on the thistle, and they were seen to present the same scene. I always noticed that the arrival of the ants and the stroking of their antennæ preceded the evacuation of the liquid, and that the attitude of the plant-lice, with their heads directed downward, seemed to be assumed for this very purpose. I witnessed this remarkable procedure thousands and thousands of times; it was always employed by the ants with the same success whenever they wished to obtain food from the