It

two pairs. Another British species, which is said to be cosmopolitan, is Metoponorthus pruinosus (Brandt). belongs to the section with two pairs of tracheæ. Mr. Whyınper found it in Ecuador at a height of 13,300 feet above the sea, highest soaring of the Isopoda. Dollfus describes Metoponorthus Barroisi, 1889, from the Azores, and in 1892 two new species from Palestine.

Rhyscotus, Budde-Lund, 1885, has a single species, Rhyscotus turgifrons from the West Indies.

Leptotrichus, Budde-Lund, 1879, with fine hairs,' receives four species, to which Dollfus doubtfully adds his Syrian Porcellio pulchellus. He also records Leptotrichus tauricus, Budde-Lund, from the Mount of Olives.

Bathytropa, Budde-Lund, 1879, 'with deep haunts,' has two or three species.

Lucasius, Kinahan, 1859, is reinstated by M. Eugène Simon in 1885, to receive not only the Algerian Porcellio myrmecophilus, Lucas, for which it was instituted by Kinahan, but also three other species, pallidus, tardus, and pauper, named by Budde-Lund and by him referred to Porcellio. In 1890 Dollfus likewise remarks that Lucasius ought to be extended to a whole group of the ancient genus Porcellio, formed of ant-loving species, with characters morphological and biological near to those of Platyarthrus. He describes a new species Lucasius hirtus from Marseilles, and from the same district records Lucasius pallidus (Budde-Lund).

Chavesia, Dollfus, 1889, agrees with the preceding genera in having the flagellum of the second antennæ two-jointed, but with Armadilloniscus more nearly in the structure of the uropods. Chavesia costulata, Dollfus, is from the Azores. By an obvious misprint the description assigns the two-jointed flagellum to the first antennæ.

Platyarthrus, Brandt, 1833, has the body broad and flattened, no eyes, the flagellum of the second antennæ small, with its first joint inconspicuous. Platyarthrus Hoffmannseggii, Brandt, appears to be met with almost all over Europe, but never except in ants' nests. In addition to the English localities named by Bate and Westwood,

there may be mentioned Tenby, Ventnor, Tunbridge Wells, and Cissbury Camp near Worthing. Probably the little white slow-moving woodlouse finds its food in the formicarium, and pays for its board and lodging by acting as a scavenger. The ants when disturbed leave their guests to shift for themselves, and do not carry them off along with their own helpless larvæ. There are two other species of the genus, one of them a cavedweller.

Oniscus, Linnæus, 1767, occupies now a humble and limited position in the sub-order of the Isopoda, the whole of which it at one time embraced. Oniscus asellus, auctorum, is very common throughout Europe and in North America. There seems no reason to relinquish the specific name in favour of the later Oniscus murarius of Cuvier. The other species found in Great Britain, Oniscus fossor, Koch, is smaller, duller in colour, and more closely tuberculate. Oniscus Simonii, Budde-Lund, occurs in the South of France; Oniscus punctatus, G. M. Thomson, in New Zealand.

Philoscia, Latreille, 1804, to be pronounced Philoskia, and meaning a lover of shade,' has a name which would probably be appropriate to every genus in this sub-order. Budde-Lund gives the description of twenty species, with the names of three more. That which he describes as Philoscia longicornis is in fact Philoscia Couchii, Kinahan, a maritime species which he elsewhere supposes possibly to belong to Ligidium. Besides inhabiting the coasts of Devon and Cornwall, Philoscia Couchii appears to be found all round the Mediterranean. It runs fast and is very shy. The body is smooth as in the rest of the genus, and the integument is not very firm. Philoscia muscorum (Scopoli) is very common in the South of England and in many parts of Europe. Budde-Lund says that Leach was mistaken in saying that it had been found in Sweden. Philoscia pubescens (Dana) is found in New Zealand. Philoscia pulchella, Budde-Lund, is identified by Dollfus with his own earlier described Philoscia elongata. Dollfus also establishes Philoscia corsica, 1888, inhabiting the

AN ANOMALOUS JOINT

431

mountain forests of Corsica, a 'beautiful species, hitherto the largest of the genus,' Philoscia celleria, 1884, from Marseilles, and Philoscia anomala, 1890, a species brought by the Challenger from Valparaiso and Juan Fernandez, in which the adult male, perhaps in the marital stage, has the fifth joint (tarsus) of the fourth pair of feet developed into an almost circular disc. Dollfus remarks that Oniscus bucculentus, Nicolet, is really a Philoscia, which also exhibits this apparently anomalous and perhaps temporary character.

Alloniscus, Dana, 1854, has seven species described by Budde-Lund, of which the first is Alloniscus perconvexus, Dana, from California; the last is Alloniscus cornutus, Budde-Lund, strongly exhibiting the generic character in having the lateral frontal processes of the head large, narrow, prominent, subconical, and the middle of the front very tumid. In addition to the seven species, Rhinoryctes mirabilis, Stuxberg, 1875, from California, is said certainly to belong to Alloniscus.

Lyprobius, Budde-Lund, 1875, receives three species. Scyphax, Dana, 1852, has the species setiger, BuddeLund, from New Caledonia, and ornatus, Dana, from a sandy beach of New Zealand.

Deto, Guérin, 1834, is said to be very near Oniscus. Deto echinata, Guérin, and Deto acinosa, Budde-Lund, are both African species.

Armadilloniscus, Uljanin, 1875, has the flagellum of the second antennæ four-jointed. Budde-Lund, besides describing three species, incorporates with them Actoniscus ellipticus, Harger, and with hesitation also Actacia euchroa, Dana, and Actæcia Aucklandiæ, G. M. Thomson, both from New Zealand. Dollfus adds Armadilloniscus tuberculatus, 1889, from the Azores. In this genus the uropods are flattened and scarcely produced, the peduncle large, and the outer branch short, just reaching the apex of the peduncle.

Scleropactes, Budde-Lund, 1885, has the body very convex and contractile into a globular form, the opercular branch in all the pleopods tracheal, and the uropods toler

ably produced with a short outer branch. The three species described by Budde-Lund are from Peru.

Budde-Lund mentions Acanthoniscus spiniger, White, from Jamaica, and Ouracharus caudatus, White, of unknown habitat, but confesses himself unable to determine their position among the Onisci.

Stenoniscus, Aubert and Dollfus, 1890, does not fall into any of the groups established or recognised by Budde-Lund. It has a very narrow, hairy body, the hind margin in the front segments of the peræon not sinuous, and but little in the hinder ones, the side-plates distinct in all but the first segment; the first segment of the pleon not visible, the others dorsally tubercled, the large rounded terminal segment completely covering the uropods, which have a broad short peduncle, the outer branch apical, minute, the inner longer than the outer, attached high up on the peduncle. The second antennæ are very short, with a two-jointed flagellum. The single species is Stenoniscus pleonalis, Aubert and Dollfus, from the sea-coast, Marseilles.

Family 5.-Armadillidida.

The body is very convex, contractile into a globular form. The head is sometimes broader than long, flanked by the first segment of the peræon; the face is nearly perpendicular. There is no demarcation of the sides of the head. The clypeus is perpendicular. The young quit the mother with all seven segments of the peræon developed. The trunk-feet are rather short. The uropods are short, flattened, not reaching beyond the postero-lateral points of the two terminal segments of the pleon. The foregoing characters distinguish this family from the Oniscidæ, the other characters of which are common to both families and need not be repeated.

Of the generic names in this family the nearest in age to the preoccupied Armadillo are Cubaris and Armadillidium. Of these two, which are contemporaries, it seemed reasonable and convenient, for the sake of old association, to give the preference to Armadillidium in forming the name of the family.

THE ARMADILLIDIDE

433

Budde-Lund gives in effect the following 'Conspectus Generum:'

I. Outer branch of the uropods small or minute, rather smooth.

A. Terminal segment not shorter than the uropods.

a. Terminal segment apically truncate.
1. Cubaris. 2. Eubelum. 3. Pseudarmadillo.
b. Terminal segment apically acuminate.
4. Peryscyphis.

B. Terminal segment shorter than the uropods,
broadly triangular.

5. Sphæroniscus. 6. Cylloma.

II. Outer branch of the uropods large, flattened, lamellar. 7. Eluma. 8. Armadillidium.

Cubaris, Brandt, 1833, appears to be the earliest synonym of Armadillo, Latreille, 1804, which is not only preoccupied, but has had the type-species removed to another genus. Under the name Armadillo, Budde-Lund gives descriptions of thirty-six species, and the names of some twenty-four more. Only two or three of this host are found in Europe, seven belong to South Africa, five to South America, a great number occur in the islands and coasts of the Pacific, while from the whole mainland of Asia not one is known, except Cubaris officinalis (Desmarest), a species found all round the Mediterranean. Dollfus adds Armadillo melanurus, 1887, Armadillo javanensis, 1889, Armadillo trifolium, 1890, and Armadillo albomarginatus, 1892. In this genus the flagellum of the second antennæ is two-jointed.

Eubelum, Budde-Lund, 1885, has the flagellum of the second antennæ three-jointed, the last segment of the pleon in the middle squarely produced. As in the preceding genus the opercular branch in all the pleopods is tracheal. The only species is Eubelum lubricum, from South Africa.

Pseudarmadillo, de Saussure, 1858, has very small, un