make them excrete the honey-dew and know exactly where to expect the evacuated liquid.

3. The ants protect the aphids. Several observers have seen the ants driving away predatory insects. Büsgen's observations on Chrysopa larvæ are cited above (p. 345). Ferton (1890), speaking of Tapinoma erraticum and its aphids, says: "While observing the aphidhunting Hymenoptera in their attacks on their prey, I was impressed with the jealous surveillance of the ants, and the protracted manœuvres of the hunters in deceiving these guardians. Cemonus unicolor Fabr. and Pemphredon insigne V. d. L., which I was especially able to follow, showed by their detours and subterfuges that their real enemy is not the aphid, but the ant which protects it." Indeed, the fierce watch fulness of Formica sanguinea or F. rufa must be apparent to any observer who disturbs these ants while they are attending their aphids. The former at once open their mandibles and rush at the intruder and the latter throw back their heads, sit up with the tips of their gasters directed forward and discharge volleys of formic acid in the direction whence they are threatened. Belt has observed the workers of Pheidole protecting their membracids in a similar manner.

4. Many aphidicolous ants, when disturbed, at once seize and carry their charges in their mandibles to a place of safety, showing very plainly their sense of ownership and interest in these helpless creatures.

5. This is also exhibited by all ants that harbor root-aphids and root-coccids in their nests. Not only are these insects kept in confinement by the ants, but they are placed by them on the roots. In order to do this the ants remove the earth from the surfaces of the roots and construct galleries and chambers around them so that the Homoptera may have easy access to their food and even move about at will.

6. Many aphidicolous and coccidicolous ants, as was shown in a former chapter (pp. 223, 224) construct, often at some distance from their nests, little closed pavilions or sheds of earth, carton or silk, as a protection for their cattle and for themselves (Figs. 205-209). This singular habit may be merely a more recent development from the older and more general habit of excavating tunnels and chambers about roots and subterranean stems.

7. The solicitude of the ants not only envelops the adult aphids and coccids, but extends also to their eggs and young. Numerous observers (Huber (1810), Lubbock (1896), Lichtenstein (1870, 1877-'80), Del Guercio, Forbes (1894), Weed (1891), Shouteden (1902), Mordwilko (1907), Webster (1907), and others) have observed ants (species of Lasius) in the autumn collecting and storing aphid eggs in the chambers of their nests, caring for them through the winter and in the spring

placing the recently hatched plant-lice on the stems and roots of the plants. Forbes, Weed and Webster have shown that our common L. americanus is thus instrumental in rearing and disseminating throughout the fields a root aphid (Aphis maidiradicis) which is very injurious to Indian corn. Webster gives the following account of his extended observations on the relations of these insects to each other: "Now, taking up the life history of the root-aphis, we find eggs in the fall, it is true, but only in the burrow of and attended by these ants. If there are eggs, egg-laying females, or males elsewhere they have yet to be discovered. The ants care for these eggs throughout the winter, shifting them about, according to Forbes, as they do their own young, to accommodate them to changes of weather and moisture. In spring, the young, as soon as they hatch from these eggs, are transferred by the ants to the roots of young fox-tail grass, smart-weed, and even ragweed. The young are carried out to pasture, as it were, during fair weather, but in bad weather, or on cold nights, they are taken back to the burrows of the ants. The plants just mentioned are the ones that push up early in spring in last year's corn lands, and especially in fields that have been plowed and allowed to stand untouched for a week or so. Usually the farmer plows his ground in spring and pays little attention to this early growth of weeds and grass, as he can generally dispose of it as soon as he begins to cultivate the corn, although this is not until the rows of young plants can be followed by the eye across an ordinary field. As soon, however, as the corn plants begin to show above ground the ants not only transfer the young root-aphids from the burrows to the roots of corn, but they will also remove them from the roots of grass and weeds and recolonize them on the roots of young corn. Now these young aphids are all females and within a few days they begin to give birth to young, also all females; these, too, are cared for by the ants, which place them on the freshest and most tender rootlets. This procedure goes on about the roots of corn throughout the spring and summer. Forbes has found that under the most favorable artificial conditions there may be as many as sixteen generations between April and October, ten of which may coexist. at the same time. It is hardly probable, however, that so many generations can exist under ordinary field conditions; nevertheless, it may be rightly inferred from this that the multiplication of the species is enorThese ants not only transfer the root-aphids from one root to another of the same plant, but will carry them from one plant to another a considerable distance away. In the spring of 1887 the writer placed a number of flower pots containing young growing uninfested corn plants between rows of infested hills of corn in the field. The corn

mous.

in the infested hills was then pulled up, exposing the roots on which the aphids were clustered. The little brown ants at once began to carry the aphids to new quarters, and the next day the latter, some of them full grown, were abundant on the roots of the corn in the pots, although there were none on them when the pots were put in place. Ants were observed over a yard away from the plants that had been uprooted, with root-aphids in their mouths, to all appearances searching for a suitable place in which to establish their charges on the roots of corn. Thus it is that from the laying of the eggs in fall to the last or egg-laying generation of the

following year this aphis is wholly dependent on the little brown ant for its existence in the cultivated fields, and the farmer can justly charge up his losses through the attacks of the root-aphis to the influences of this ant. But the matter does not terminate here, as will be seen by what follows.

I

FIG. 212. Membracid (Centrotus sp.) of Ceylon. (Green.) a, Larva from the right side; b, protruded anal segments of same.

"So long as the roots upon which the root-aphids are colonized afford an abundance of nourishment for them, all will be wingless, but as soon as the roots become tough and woody or dry out there will be a generation of both winged and wingless individuals, the former escaping from the burrows about the roots to fly to other plants, and in all probability to other fields, where they may be found on the leaves. The ants usually transfer the wingless females to more succulent roots, but seem to pay little or no attention to the winged individuals, letting these make their way out and away. But in May, 1887, the writer was able to watch some of these winged nomadic individuals in a corn field to which they had migrated and to note the results of their wanderings. A field of corn had been planted on May 18. Five days later there came a heavy rain storm that flattened the surface of the ground, which was soon encrusted by the action of the wind and sun. Four days afterwards there were freshly thrown up mounds of earth about some of the corn plants, and ants were busily engaged in and about these and running up and down over the young corn. On examining these mounds and burrows the writer was surprised to find winged root-aphids giving birth to young on the roots attended by ants. All of these young were very small, at most but a few days old. Other winged individuals were found on the leaves and even on the stems of

corn, and when any one of these was placed where the ants could find it, it was promptly captured by an ant and transported to the roots of the corn. Observation showed that as soon as the ants running about over the young corn plants found a winged aphis they made a burrow about the base of a plant, and soon domiciled the wanderer on the root under their guardianship. Then when the aphis began to give birth to young these were promptly removed to another part of the same root, or to another root close by, and there watched over by the patient and industrious ants. The same thing was observed going on about a young plant of fox-tail grass in this same corn-field."

8. Some writers have described the ants as facilitating the exit of the winged sexual generation of aphids from their nests by opening up galleries for this very purpose to the surface of the soil.

9. Others (Lichtenstein, Mordwilko) have seen ants (Lasius niger, flavus and umbratus) in the act of clipping off the wings of female aphids. Whether this is done because these organs hinder the ants from imbibing the honey-dew, or in order to prevent the escape of the aphids from the nest, or for some other reason, has not been determined.

10. The single structural adaptation which may, perhaps, have been developed through association with aphids, coccids and membracids is the greatly distensible crop of the Camponotinæ. It is quite as probable, however, that the peculiar properties of this organ may be due to the habit of collecting and retaining large quantities of nectar or other plant-secretions.

Certain Fulgorida, at least in Europe and North Africa, have intimate relations with ants, according to the observations on species of Issus, and especially of Tettigometra, contributed by Rouget (1866), Puton (1869), Lichtenstein (1870, 1880), Delpino (1872, 1875), Forel (1890, 1894), Schneider (1893), Silvestri (1903), Lesne (1905) and Torka (1905). Silvestri has studied T. impressifrons and T. costatus, which live in the nests of Tapinoma nigerrimum. The young fulgorids which, like the radicicolous aphids and coccids, obtain their food from roots and underground stems, furnish the ants with a sweet liquid. This is not excrement, however, as in the case of the other Homoptera above described, but, according to Silvestri, "a secretion from cellular glands, distributed in areas on the following segments of the body: dorsally two (i. e., one on each side of the median sagittal plane) on the submedian portion of the prothorax, two on the submedian region of the second abdominal segment and two on the sublateral portions of each abdominal segment from the third to the seventh inclusive; and ventrally, two on the lateral portions of the prothorax, two on the median portion of the third abdominal segment and two

sublaterally on each abdominal segment from the fourth to the seventh inclusive." This author believes that the larvæ of a Coccinellid beetle, Hyperaspis reppensis, which live in the Tapinoma nests and are treated by the ants as true myrmecophiles, feed on the Tettigometra. The adult beetles, however, seem to be less indulgently regarded by the ants. Torka finds that the larvæ of T. obliqua, found in great numbers about the roots of oats and rye, are assiduously attended by Formica cinerea and Lasius niger.

Perhaps further investigations will bring to light a number of cases in which Heteropterous Hemiptera are attended by ants. One such case, in fact, has been recorded, namely, a Ceylonese species of Coptosoma which according to Green (1900) is attended by Cremastogaster in the same manner as the above described Homoptera. Wasmann (1901) mentions the larvæ of Neoblissus parasitaster as living in the nests of Solenopsis geminata in Brazil, but this may be a case of true myrmecophily.

The relations of ants to the caterpillars of the Lycænid butterflies have been repeatedly described by a number of observers: Freyer (1836), Ploetz (1865), McCook (1877), W. H. Edwards (1878a, 18786, 1884), Saunders (1878), Thwaits (1881), Miskin (1883), Aurivillius (1884, 1887), Doherty (1886), Scudder (1888), de Nicéville (1888, 1889, 1890, 1900), von Aigner-Abafi (1898, 1899), Wroughton (1892), Thomann (1901), Dodd (1902a, 1902b), Green (1902), Frohawk (1904, 1906) and Viehmeyer (1907). The most comprehensive accounts have been furnished by Edwards, de Nicéville, Thomann and Viehmeyer. No less than sixty-five species, representing twenty-nine genera of Lycænidæ, are mentioned as having caterpillars that are attended by ants. According to Viehmeyer the list embraces twentythree species of Lycana alone. The larvæ of this and the allied genera are somewhat depressed with rounded anterior and posterior ends and with the tense and highly sensitive skin covered with short, sparse hairs (Figs. 210 and 211). As Guenée long ago (1867) showed for Lycana bætica, these caterpillars possess three peculiar organs, one an unpaired gland (a, x) in the middorsal line of the antepenultimate, or eleventh segment, and a pair of short protrusible tentacles on the dorsolateral portions of the penultimate, or twelfth segment (b). The median gland has the form of a sac or papilla which can be protruded through a transverse, mouth-shaped slit, and each of the tentacles is fringed at its tip with a dense circlet of stiff, finely plumose hairs (Fig. 211, d, b'). The ants caress the posterior, somewhat flattened extremity of the caterpillar with their antennæ as they do that of the plant-louse, and the caterpillar responds by emitting from the