size and greater whiteness. The coarctate skin, when deserted, retains its original form almost intact. The third larva is rather active, and burrows about in the ground; but while there seems to be no reason why it should not feed, nourishment is not at all essential, and all my specimens have, in the course of a few days, transformed to the true pupa without feeding. In the transformation to pupa (Pl. 1., Fig. 9) the third larval skin is worked into a wrinkled mass behind, as is also the skin of the true pupa when shed. The pupa state lasts but five or six days, and before the wings of the imago are fully expanded, or the abdomen contracted, the general aspect of Epicauta forcibly recalls the mature Henous.

Like all parasitic1 insects that nourish on a limited amount of food and possess no power to secure more, the blister-beetles vary greatly in individual size in the same species, and the larvæ have the power of accommodating their life to circumstances, and of assuming the coarctate larval form earlier or later according to the size of the egg-mass which they infest. I have had some interesting illustrations of this in my experiments with them. In an average sized egg-pod of the differential locust, however, there are more than enough eggs to nourish the largest specimen of E. vittata, and a few are usually left untouched.

The period of growth, from the first feeding to the coarctate larva, averages, as will be gathered from the foregoing, about a month; yet in the month of September, out-doors, under screens where I have had the differential locust oviposit for the experiment, I have known the full larval growth of E. vittata to occupy but 24 days. As this species occurs in the beetle state as early as June in the latitude of St. Louis and as late as October, there are possibly two annual generations here and farther south.

Larval Habits of Macrobasis and Henous.-The characteristics of the triungulins of the blister-beetles, represented by Epicauta and Henous, are remarkably similar, and point to unity of habit. The same holds true of the characters of the second, coarctate and third larva and of the pupa of Epicauta and Macrobasis. They are precisely alike; so that, while appreciable differences may be found in the triungulins, it is doubtful whether the subse

1An insect is not properly parasitic that simply feeds on eggs, but the term is permissible and even necessary to characterize and distinguish those species which develop within and are confined to a locust egg-pod, from the predaceous species that are not confined but pass from one pod to another.

quent developmental stages will indicate specific or even generic differences in species of similar size in these three genera.

That the eggs of Epicauta may exceptionally hybernate is possible, but, from their delicate nature, improbable. That the triungulins frequently do so there can be no doubt, especially in species like the Black Blister-beetle, which is found on the flowers of Solidago, Eupatorium, etc., till the end of October, and continues laying till frost. I have at the present time (November) many of these last that are quietly huddled together, and, with winter temperature, will doubtless remain so; while others have worked in between the locust eggs, there evidently to remain without feeding till spring opens. I have also found as many as five triungulins of this species curled up in the deep red mucous matter that surrounds the eggs of Edipoda phanecoptera-all numb and torpid, and evidently hybernating.

Conclusion.-From the foregoing history of our commoner blister-beetles, it is clear that while they pass through the curious hypermetamorphoses so characteristic of the family, and have many other features in common, yet Epicauta and Macrobasis differ in many important respects from Meloë and Sitaris, the only genera hitherto fully known biologically. To resume what is known of the larval habits of the family, we have:

1st-The small, smooth, unarmed, tapering triungulin of the prolific Sitaris, with the thoracic joints subequal, with strong articulating, tarsal claws on the stout-thighed but spineless1 legs, and, in addition, a caudal spinning apparatus. The mandibles scarcely extend beyond the labrum; the creature seeks the light, and is admirably adapted to adhering to bees but not to burrowing in the ground. The second larva is mellivorous, and the transformations from the coarctate larval stage all take place within the unrent larval skin.—We have:

2d-The more spinous and larger triungulin of the still more prolific Meloë, with long caudal setæ, but otherwise closely resembling that of Sitaris in the femoral, tarsal and trophial characters, in the subequal thoracic joints, in the unarmed tibiæ, and in the instinctive love of light and fondness for fastening to bees. The second larva is also mellivorous, but the later transformations take place in the rent and partly shed skins of the second and coarctate larvæ.—We have :

1 The larva of S. humeralis appears to differ from that of S. colletis in having hairs on the femora and tibiæ.

3d. The larger and much more spinous triungulins of the less prolific Epicauta, Macrobasis, and Henous; with unequal thoracic joints, powerful mandibles and maxillæ, shortened labrum, slender femora, well-armed tibiæ, slender, spine-like, less perfect tarsal claws-combined with an instinctive love of darkness and tendency to burrow and hide in the ground. The second larva takes the same food as the first, its skin is almost entirely cast from the coarctate larva, while the subsequent changes are independent and entirely free of the shell of this last.

Larval Habits of Cantharis.-The question naturally arises here, whether Cantharis, in its larval habits, will most agree with Meloë and Sitaris or with Epicanta. The triungulin, except in becoming almost black, has much in common with Meloë, in the subequal thoracic joints, and the long antennæ; also in its habit, observed by Lichtenstein, of fastening to bees. The fact that it can nourish on honey, though it does not appear to do so freely, would also indicate that it breeds in the nests of solitary bees. Nevertheless, in the slender thighs and the caudal and abdominal characters it agrees more nearly with Epicauta, and in the stage following the first molt the legs are still quite long and the general aspect much like the carabidoid stage of that genus. I should not be surprised, therefore, if Cantharis also nourished on locust eggs. What is Known of the Larval Habits of other Meloid Genera.— Mylabris Fabr. (nec Geoff.), according to V.-Mayet, is much less prolific than any Meloids so far observed. The egg is 2.5 mm. long and is 1⁄2 as wide, with a tolerably thick shell and the embryo more fully bent within it. The triungulin has many of the characters of Epicauta, judging from the published description (Ann. Soc. Ent. de Fr., 1876, p. cxcvi.), which is, however, not sufficiently detailed as to the trophi. I doubt not that the genus will be found to infest locust eggs.

Horia Fabr., from what little is known of it, would seem to have a similar partial parasitism to Meloë, but on carpenter bees. Tetraonyx Latr., was found by Guérin-Meneville in places frequented by bumble-bees.

The eggs of Apalus Fabr., as well as its triungulin, are said to resemble those of Meloë.

Zonitis Fabr., is known to develop in the cells of Osmia and Anthidium, and to have a coarctate larva much like that of Sitaris.

1 Since this was published M. Lichtenstein writes me: "They only fasten to bees when enclosed in tubes with them. At liberty they do not climb flowers, like Meloë, in search of bees, but run quickly on the soil, just as you describe for Epicauta."

AS

THE SIRENIA.

BY ARTHUR E. BROWN.

S the name of this curious order of aquatic mammals was suggested, without doubt, by the probability that they, in great measure, gave origin to the ancient myths regarding the existence of sirens or mermaids, it may not be inappropriate to take a first look at the animal as it appeared to a sailor of Capt. Weddell's Expedition towards the South Pole, about the close of the first quarter of the present century. "The sailor had lain down and about ten o'clock he heard a noise resembling human cries; and as daylight in these latitudes never disappears at this season, he rose and looked round, but on seeing no person he returned to bed; presently he heard the noise again, rose a second time, but still saw nothing. Conceiving, however, the possibility of a boat being upset, and that some of the crew might be clinging to some detached rocks, he walked along the beach a few steps and heard the noise more distinctly, but in a musical strain. Upon searching round he saw an object lying on a rock a dozen yards from the shore at which he was somewhat frightened. The face and shoulders appeared of human form and of a reddish color; over the shoulders hung long green hair; the tail resembled that of the seal, but the extremities of the arms he could not see distinctly. The creature continued to make a musical noise while he gazed about two minutes, and on perceiving him it disappeared in an instant."

Notwithstanding the fact that the sailor testified his belief in his story by making a cross in the sand and kissing it, we who look on the manatee and dugong, with minds cleared from a delusive belief in mermaids and mermen, are tempted to agree with Capt. Weddell in the last clause of which he makes use in relating the circumstance: "I concluded he must really have seen the animal he described, or that it must have been the effects of a disturbed imagination."

Naturalists, however, have had almost as much trouble, in times. past, to determine where these animals structurally belong, as the old explorers had to decide whether they were forms of earthly or supernatural origin. They were classed in the old order Pachydermata on account of their points of resemblance to the elephant; then they were put with the whale and the porpoise in

the Cetacea, and they are now generally regarded as an independent order, known as Sirenia, allied more closely with the Ungulates than with any other living mammals; they-the vegetable eaters of the sea-being doubtless descended from some form of terrestrial Herbivora, as the Seals, Whales, etc., the aquatic Carnivores-probably claim parentage among the terrestrial Car

nivora.

The earliest Sirenian known to Palæontology is the Ilalitherium, found in the Miocene formation of Central and Southern Europe. Those now existing are separated into two genera,-Manatus and Halicore, and, generally speaking, there is a strong resemblance between all members of the group.

They all possess a cylindrical body tapering towards each end from a point just in advance of the middle; the tail is flat and horizontal; the hind limbs are entirely wanting; the fore limbs are placed far forward on the breast and are covered with integument like the rest of the body; the nipples are two in number, and, as in the human species, are situated on the breast; the neck is thick and is not separated by a well drawn line of demarcation from either the head or the body; the head is rather small and round, the lips being thick and fleshy; the eyes are round and of small size and the nostrils, situated on top of the extreme end of the snout, are closed with a valve; the presence of an ear is only to be recognized by a sort of crease in the skin; the hide, in appearance and toughness, much resembles galvanized rubber and is covered very scantily with stiff bristles an inch or two long, these are so sparsely scattered, however, that it requires a close inspection, almost, to see that they are there,-around the lips and mouth they are longer and more plenty and are deserving almost of the name of whiskers. In these characters all the Sirenians agree.

The two genera composing the order are, however, recognizable at a glance, and the species into which they are divided, though closely allied, yet present characters which are held to be sufficiently distinctive for the recognition of two species of dugong and three of manatce.

The one genus, Manatus, is found altogether in the waters of the Atlantic Ocean, with the single exception of a small part of the East coast of Africa, north of the Cape of Good Hope, while the other genus, Halicore, is strictly confined to the South Pacific and Indian Oceans. No species are found in the Northern