tory parasite is not a bed of roses. The amazons have indeed acquired brilliant military instincts, but if these ants were capable of reflection they might occasionally regret having abandoned the quiet pastoral life which their ancestors probably led.

(c) Polyergus lucidus (Figs. 270 and 271).—This is the largest, handsomest and most graceful of our amazons, and even surpasses the European form in its brilliant red coloring and gleaming surface. It may be called the shining amazon. Owing to its wide distribution in the Eastern States it has been known for some time. Mrs. Treat (1877) and McCook (1880b) saw its colonies, but Burrill (1908) is the only author who has described one of its forays. In the Atlantic States it is very rare and sporadic. During the past five years I have seen only four of its colonies in New York and New Jersey, and these were at great distances from one another. It is more frequently met with on the warm eastern slopes of the Rocky Mountains in Colorado, where, unlike breviceps, it occurs only at lower elevations. It is the furthest removed of all of our subspecies from the European type in its smooth surface, in the coloring of its queens, which, in eastern colonies, at least, have the head and thorax nearly black, in the small size of its communities and the character of its slaves. These are not members of the fusca, but of the pallide-fulva group of Formica, and are represented by schaufussi, or the closely allied incerta or nitidiventris. The very small size of the colonies of these ants may account for the same peculiarity in lucidus. The slave species makes obscure crater nests in sunny, open pastures and such places are therefore also the home of the shining amazons. The ratio of these to slaves in the mixed colonies is about 1:5 or 6. It is an interesting fact that lucidus resembles its slaves in having a smooth, shining surface, a slender, elegant stature and long legs, whereas breviceps and bicolor resemble the fusca forms which they enslave, in having a more pilose and pubescent surface, more thickset stature and shorter legs. These resemblances may therefore be regarded as mimetic.

Near my former home in Bronxville, N. Y., there was an unusually fine lucidus-incerta colony, which I had under observation for five years. During four years this colony produced numbers of males and females, both winged and ergatoid, and the winged females lingered for weeks in the nest without deälation. The first week of April 1908, I found the whole community with its larvæ and mother queen enjoying the spring warmth in the superficial galleries just under the large flat stone with which I covered the nest in September 1903. I captured the queen and part of the colony and transferred them to an artificial nest. August 9 I again visited the nest, and to my

surprise, found it teeming with several hundred males clinging to the lower surface of the stone, but with no winged or deälated females. Besides the males I found only a single large ergatoid female, several dozen workers and slaves, and half a dozen cocoons enclosing nearly mature male pupæ. Without doubt, the ergatoid had usurped the rôle of the mother queen and, being unfertilized, had produced only male offspring. The comparatively small number of slaves had been able to rear an enormous number of these little creatures, although the absence of incerta pupa in the nest indicated that the Polyergus

FIG. 271. Polyergus lucidus. (Original.) a, Worker in profile; b, head from above.

workers had made no forays during the summer of 1908. The following is a description of one of the forays made by this colony July 31, 1904, while it was still in a normal condition:

On reaching the nest at 2.20 P. M. I found the lucidus pouring out of it in numbers. They ran about on the stones and surrounding soil till 2.37, when a troop of nearly 200 had congregated and began to move away from the nest, slowly at first and then with feverish paces. At 2.45 they reached, by a direct path through the grass, an obscure crater nest of incerta, situated some fifteen meters from their own. They at once poured into the opening, slaughtering or putting to flight the incerta that were loitering about or issuing from the galleries. One minute later (at 2.46) the first lucidus emerged with a cocoon. Then followed a stream of these ants, each similarly laden, and started for home. Several were unable to secure sound pupæ, but grabbed up empty cocoons and pupal exuviæ and fell in line. Some also brought out recently hatched and callow incerta and slaughtered them on the nest crater. The last cocoon was brought out at 3.13 and a few moments later the last amazon left the nest and joined the returning troop. During the pillage some of the incerta endeavored to defend their nest,

but were promptly dispatched by the lucidus. The corpses were dragged away by a lot of Myrmica sabuleti that had their nest about the roots of a plant within 30 cm. of the incerta nest. By 3 P. M. the last lucidus had disappeared into her nest. The whole expedition therefore consumed only forty minutes.

During the summer of 1902 I found near Rockford, Ill., a fine lucidus-nitidiventris colony. Thirty of the amazons were transferred to an artificial nest furnished with a wet sponge and a dish of honey, for the purpose of studying their behavior when isolated from their slaves. In the course of a few days the ants were famished and kept vainly begging one another for food. They often licked the water from the surface of the sponge and two that had accidentally stumbled into the honey head foremost, so that their tongues were brought in contact with it, lapped it up with avidity. They soon began to die of hunger and when only sixteen of them survived I wished to see whether they would adopt alien workers of subsericea and nitidiventris taken from a garden far removed from any lucidus colony. At two o'clock one afternoon three of the former (A', B', C') and three of the latter (A, B, C) were placed in the nest. They began to run about in great dismay, especially when they happened to touch one of the amazons with their antennæ. From the first, however, the subsericea seemed to be much more frightened and to irritate the amazons more than did the nitidiventris. One large lucidus that had lost her right antenna and right tarsus seemed to be in a particularly vicious frame of mind, possibly because she had been spending much of the day trying to comb an imaginary antenna with an imaginary strigil and had repeatedly tumbled over while attempting this feat. This ant pounced on A' and B' like a cat and killed them in quick succession. She pierced the gaster of A' with her sharp jaws till its contents flowed out on the floor of the nest. B' she pierced through the thorax. For some time the surviving subsericea (C') succeeded in evading the amazons, which in the meantime had worked themselves up into such a fury that they even attacked one another. I saw one of them grab a sister worker by the neck and hold on for three quarters of an hour. The amazons, though visibly irritated by the nitidiventris, did not seize them by the body, but only by the legs and antennæ. The smallest individual (A) lost its left middle leg in such an encounter. At this point the observations were interrupted.

At 9 P. M. both A and B were found dead, and C and C' were running about. C' skulked in the corners of the nest, but C was seen to walk up to one of the amazons, protrude her tongue, and feed the

famished creature for several minutes, then go to the food chamber, take a draught of honey, return and feed a number of them in succession. For an hour she thus moved back and forth between the amazons and the honey till all had been fed. At 8 A. M. on the following morning the amazons were huddled together on the sponge, as if asleep, and in their midst was the nitidiventris, C, also resting peacefully. The subsericea, C', still alive but distrustful, was lurking in the furthest corner. I tapped the nest gently to arouse its inmates. C immediately ran into the food chamber, imbibed a lot of honey, returned and began to feed the lucidus, like a solicitous mother who wakens early and sets about getting breakfast for a large family. The friendly relations between C and the amazons continued throughout the day. Early the following morning C' was found dead in a corner and C was ministering to the amazons. Before noon, however, I found her lan

guidly dragging her body about the food chamber. Her head had been pierced by one of the stupid and ungrateful lucidus. She died in a few hours.

At 3 P. M. twenty large subsericea workers, with a number of larvæ and pupæ, were placed in the light chamber. The sixteen amazons entered the chamber and began to attack the black intruders in a perfect frenzy of valor. To my surprise the subsericea stood their ground, took the offensive and were soon driving the amazons around the nest like a herd of sheep. They seized the lucidus by the legs and antennæ,

showered them with formic acid, mauled them about, gnawed off their legs and left them in a pitiable plight. This victory for the subsericea was not only a surprise, but coming so soon after the death of the selfsacrificing nitidiventris, made me feel much as I felt when a boy on reading of the death of the suitors in the Odyssey.

These observations on lucidus, with many others which space forbids relating, show that this subspecies is much more belligerent and of a more vicious disposition than breviceps or bicolor. This is the more surprising because the ants which it enslaves are more cowardly and docile than the slaves of the other subspecies. The behavior of the ministering nitidiventris also shows that this form and not subsericea is the natural slave of lucidus.

3. The Founding of Amazon Colonies.-Most authors have inferred from the absence of the domestic instincts in the amazons that the queens of these ants would be unable to establish a formicary without the aid of alien workers. Forel and Wasmann have therefore insisted that the rufescens queen must be adopted by a band of fusca, and they have published several observations which go to show that such an adoption can be rather easily brought about in artificial nests. These observations have been recently confirmed and extended by Viehmeyer (1908). In this respect Polyergus rufescens resembles the temporary parasites. Several experiments in which I introduced artificially deälated queens of lucidus into nests containing incerta workers with their brood gave rather conflicting results. In some cases the lucidus queens behaved like sanguinea queens under similar conditions, to the extent of killing the alien workers, but they paid absolutely no attention to the brood. In other cases they were more passive and conciliatory, but equally indifferent to the incerta cocoons. It will be necessary, therefore to study this question further before making definite statements in regard to the method employed by our American amazons in establishing colonies. But even if the method of rufescens should be found to obtain also in our subspecies, we should not be justified in deriving it from that of the temporary social parasites, for we might conceive it to have arisen secondarily by involution or degeneration from that employed by sanguinea.1

'Professor Emery, in a paper just received (Nuove Osservazioni ed Esperimenti sulla Formica Amazzone. Rend. Sess. R. Accad. Sci. Inst. Bologna, 1909, pp. 31-36), records an experiment which goes a long way towards solving the problem here considered. July 13 he placed a deälated P. rufescens queen in a Janet nest containing a F. fusca queen, fourteen workers and a few pupæ. The rufescens, on being attacked by the workers, offered no resistence, but showed great interest in the fusca queen, who received her amicably. By July