The perplexing question of the affinities of the Brachiopoda has given rise to much discussion, and great difference of opinion, especially with regard to their relationship to the group of worms. of worms. Now, although I do not admit the Brachiopoda to be worms, they may, as well as the Mollusca and some other groups of invertebrates, have originally diverged from an ancestral vermiform stem, such as the remarkable worm-like mollusk Neomenia would denote. In a recent paper on the development of Argiope or Cistella, Mr. A. E. Shipley observes, and, I think, with justice, that the Brachiopoda and Polyzoa are not so closely united as to form a natural phylum; and he adds, "I should propose to follow Gegenbaur in making a primary class of the Brachiopoda, and though in their development and adult structure they are widely separated from both Vermes and Mollusca, of the two classes I would place them nearer to the former class than to the latter"*. Prof. Huxley + says:-" All known Polyzoa are compound animals, that is to say, the product of every ovum gives rise, by gemmation, to great assemblages of partially independent organisms, or zooids. The Brachiopoda, on the contrary, are all simple, the product of each ovum not giving rise to others by gemmation. All the Brachiopoda possess a bivalve shell-a shell composed of two, more or less horny, or calcified, pieces, which are capable of a certain range of motion on one another, and are very commonly articulated together by teeth and sockets." The shell, the pallial lobes, the intestine, the nerves, and the atrial system, afford characters amply sufficient to define the class. In this view of Prof. Huxley I entirely concur.

As many species of Brachiopoda live at considerable depths, it is not surprising that so small a number should have been known to early conchologists, and that for many years they should have been such great rarities in conchological collections. The numerous well-conducted dredging expeditions have, however, brought to light a large number of forms that were not previously known, and we may constantly expect to add to the number of species as dredging operations extend to regions not yet explored. It has been ascertained beyond doubt that Brachiopoda are much localized, and that where they occur they are generally abundant. It has also been found that the range in depth of one and the same species is often very variable, that abyssal forms have generally a very thin shell, and that species living at a great depth have a much greater geographical range, and are not nearly so localized as those species that live in shallow waters.

The study of the species brought home by the Challenger' Expedition, which I was privileged to examine and describe, has revealed much valuable information with respect to the bathymetrical and geographical distribution of many species. The greatest depth at which a recent species of the class has been found alive was 2900 fathoms. A number of forms inhabit and prefer rocky and stony parts of the bottom, or are attached to corals, and are therefore more difficult to obtain.

It is necessary briefly to refer to the difficult question of classification, upon which many different opinions have been entertained. In company with a larger number of

* "On the Structure and Development of Argiope." Mittheilungen aus der zool. Station zu Neapel, Band iv. Heft 4, p. 516 (1883).

† An Introduction to the Classification of Animals, p. 27 (1869).

malacologists and palæontologists, I have considered the interior skeleton that supports the labial appendages as a classificatory character that could be advantageously made use of, and have consequently grouped the recent species into the two great divisions Arthropomata, Owen (=Clistenterata, King), and Lyopomata, Owen (=Tretenterata, King), and into six families, as follows:

In 1884, M. E. E. Deslongchamps proposed a new scheme of classification for the Terebratulidæ, in which he objected to any arrangement based on either the exterior shape of the shell or of the supports of the labial appendage.

His first group includes the different forms in which the calcified brachial apparatus or loop does not undergo any important modifications from its first origin up to the adult condition. To the characters drawn from the brachial apparatus or loop is added that of the presence of spicula, more or less complicated, which occupy in the mantle all the parts connected with the organs of circulation (arteries, veins or veiny sinuses, &c.), the labial appendages and cirri which accompany them. In this group he places the recent genera Liothyris, Terebratulina, Megerlia, Kraussina, and Platydia.

In his second group he unites those forms in which the brachial apparatus or loop undergoes numerous transformations from the embryo up to the adult condition, and which have been distinguished by the names of Platydiform, Magadiform, and Megerli form stages. The mantle in this group is not provided with those calcareous spicula which occur so constantly in the first group. The colour also of the dried animal is yellowish, whilst in the first group the hue or tint of the dried arms and of the peripheric portions of the mantle are of a very characteristic dim white. This, I may, however, remark, is not always the case, as I possess specimens of the dried animal of several species of Terebratulina that are of a decided yellow colour. The group would, according to M. Deslongchamps, comprise the recent genera Waldheimia, Macandrevia, Terebratella, Laqueus, and Magasella.

While fully appreciating the importance of all characters derived from a study of the animal, I am not convinced that the temporary modifications in the shape of the loop, or the presence or absence of calcareous spicula in the mantle, &c., are indications of sufficient importance or permanence to supersede those derived from the adult shape of the calcareous lamellæ supporting the labial appendages,-characters which are often accessible, and of important assistance in distinguishing the more numerous fossil members of the group. Moreover, Mr. W. H. Dall, in describing the animal of Waldheimia floridana (Bull. Mus. Comp. Zoöl. Harvard, vol. iii. p. 16, 1871), distinctly notes the existence of a few exceedingly delicate spicula in the floor of the great sinuses. It would seem therefore that these spicula occur in a genus which M. E. E. Deslongchamps, in his proposed new classification of the Terebratulidæ, places among those forms characterized by the entire absence of spicula.

After long and searching examinations of the recent forms, I have described in this monograph about one hundred so-termed species, some varieties, and about twentyeight uncertain ones. Of course the vexed question as to what really constitutes a species remains the same, and is likely to remain so for a long time to come.

It will not be necessary to extend these introductory remarks, as all details have been fully given under each species. In conclusion I would tender my grateful thanks to the many kind friends who have in so zealous a manner supplied me with valuable information and specimens *.

ARTHROPOMATA, Owen=CLISTENTERATA, King.

Family TEREBRATULIDÆ, (Gray) emend. Davidson.

Subfamily TEREBRATULINÆ, Dall, 1870.

During the last few years a strong desire has been manifested by those palæontologists who consider an extreme subdivision of genera desirable, to separate from Terebratula proper those forms characterized by a small short loop, of which the principal stems are united anteriorly by a slightly arched lamella, and of which Liothyris vitrea may be taken as the type.

* The drawings for the Plates were made by myself, on paper; but the state of my health would not allow of my reproducing them on stone.

As there are certainly some differences observable in the two groups, and as none of the recent species would agree in the characters of their loops and in certain other particulars with the forms referred to the genus Terebratula of Llhwyd and Klein, it may perhaps be better to adopt Douville's generic name Liothyris for the species we are about to describe.

In the larger number of the recent species, such as in Liothyris vitrea, L. arctica, L. Moseleyi, L. uva, L. Bartletti, L. Wyvillii, and L. subquadrata, the connecting band of the loop is narrow, while in L. sphænoidea=cubensis it is larger.

The specific claims of Liothyris cernica, and L.? Dalli are still uncertain, only a single example of each of them having been hitherto discovered.

Very small, or scarcely any modifications in the shape of the loop have been observed; all the species have their shell minutely perforated by canals, and calcareous spicules are abundant in the mantle.

1. LIOTHYRIS VITREA, Born, sp. (Plate I. figs. 1-12.)

Anomia vitrea, Born, Testacca Musei Caes. p. 119, vignette, 1778; Linné, ed. Gmelin, p. 3347, 1788.

Gryphus vitrea, Megerle v. Mühlfeld, Berlin Mus. 1811.

Terebratula vitrea, Lamarck, An. sans Vert. vol. vii. p. 245, 1819; Payraudeau, Cat. p. 83, no. 160, 1826; G. Bronn, Italiens Tertiär-Gebilde, p. 125, 1831; Philippi, Enum. Moll. Siciliæ, vol. i. p. 95, t. 6. figs. 6-8, 1836, vol. ii. p. 66, 1844; A. Scacchi, Cat. Conch. Regni Neapolitani, p. 8, 1836; Küster, Martini & Chemnitz, Conch.-Cab. vol. vii. p. 22, tab. 2. figs. 11-13, 1813; E. Forbes, Report on the Mollusca of the Egean Sea, Brit. Assoc. Report, p. 141, 1843; D. Galvani, Illustrazione delle Conch. Foss. 1815; G. B. Sowerby, Thes. Conch. vol. i. p. 353, pl. 70. figs. 56-59, 1846; Aradas (pars) Conchiglie fossili di Gravatelli, p. 14, 1851; O. G. Costa, Fauna del regno di Napoli, p. 33, pl. i. figs. 1-3, 1851-52; Davidson, Sketch of a Class. of recent Brachiopoda, Ann. & Mag. Nat. Hist. vol. ix. p. 364, 1852, and Br. Foss. Brach. Pal. Soc. vol. i., Introduction, p. 62, fig. 23, and pl. vi. 1852; S. P. Woodward, Manual of Mollusca, p. 215, 1856; L. Reeve, Conch. Icon. pl. 3. fig. 8, 1860, and Journ. de Conch. vol. ix. p. 124, 1861; Chenu, Man. de Conch. vol. ii. p. 201, 1862; Seguenza, Atti della Soc. Italiana di Scienze Nat. vol. i. p. 17, pl. 1. figs. 1-7, 1865; H. C. Weinkauff, Die Conch. Mittelmeeres, vol. i. p. 284, 1867; Davidson, Italian Tert. Brach., Ann. & Mag. Nat. Hist. vol. vii. pl. xvii. fig. 11, 1870; Jeffreys, Ann. & Mag. Nat. Hist. 5th scr. vol. x. p. 28, 1882.

Liothyris vitrea, Douvillé, Bull. Soc. Géol. de France, 3o sér. vol. vii. 1879; E. Deslongchamps, Note sur la Classification des Térébratules, ou Etudes critiques sur les Brachiopodes, pp. 106 & 153, pl. xx. figs. 7-11, 1884.

Shell longitudinally oval or ovate, globose, widest about the middle, laterally rounded or more or less pinched in near the front, front margin nearly straight or gently rounded. Colour nearly white, surface smooth, semitransparent, glassy, marked with fine concentric lines of growth and perforated by minute canals. Dorsal valve tumidly convex, longitudinally flattened along the middle, from which the anterior lateral portions slope to the edge. Ventral valve slightly deeper than the dorsal one, longitudinally flattened along the middle; beak incurved, moderately produced and slightly overlying the umbo of the dorsal valve, obliquely truncated by a very small circular foramen with thickened margin and separated from the hinge-line by a small triangular deltidium in two pieces. Loop in the interior of the dorsal valve simple and short, attached by its

crura to the hinge-plate and not exceeding a fourth of the length of the valve, the two principal stems becoming soon united anteriorly by a transverse lamella bent upwards in the middle; no mesial septum; cardinal process small and prominent, hinge-plate disunited, four diverging grooves extending from under the cardinal process to about half the length of the valve, the central pair being the longest. In the interior of the ventral valve a similar number of grooves; muscular impressions small, situated at the bottom of the valve under the loop and in the rostral portion of the bottom of the ventral one. Animal attached by a peduncle; labial appendages united to each other by a membrane; brachial disk trilobed; central lobe elongated and spirally convoluted. Very delicate spicula form elegant star-like plates in the mantle. Length 1 inch 8 lines, breadth 1 inch 5 lines, depth 1 inch.

Hab. Abundant in the Mediterranean in depths of from 90 to 250 fathoms. Very numerous in the Bay of Naples at depths of from 100 to 300 metres. Vigo Bay, 40 fathoms (McAndrew). Dredged by Prof. Giglioli, during the Italian Expedition to the Mediterranean in 1881, at a depth of 800 fathoms (see report in the Atti del iii. Congresso Geografico Internazionale). Also for distribution of this and other species, see Jeffreys's papers "On the Mollusca of the Lightning' and 'Porcupine' Expeditions," Proc. Zool. Soc. Lond. 1878, 1879, 1881, 1882.

Obs. Liothyris vitrea is a beautiful, well-known, and abundant Mediterranean species. It varies considerably in its relative length, breadth, and degree of convexity; some specimens being quite elongated oval, 1 inch 8 lines in length by 1 inch 1 line in breadth, while other examples of the same length would have a breadth of 1 inch and 3 or 4 lines; some are much more pinched in anteriorly than others, and, lastly, some are nearly circular with an equal length and breadth.

Prof. E. Deslongchamps, in his instructive memoir on the classification of the Terebratulidæ, gives us the result of his studies in connection with the embryo of L. vitrea, as well as of its subsequent stages of development. He states, "I have been able to examine the embryo at two millimètres of size, that is to say from the first moments when the larval condition has ended and the shell has begun to be formed; its shape is then absolutely similar to that of the young of Terebratulina; the dorsal valve is rounded, and slightly convex; the ventral one shows a triangular hole, of which the summit, which does not yet show any trace of a notch, will eventually become the beak. No trace of a deltidium is to be seen on the sides of this foramen. On opening this little shell, one is at once struck by the dull white of the internal walls; and, by the aid of a lens, one recognizes granulous parts affecting a certain regularity. The same dull white condition is seen on the fragments of the labial appendages that have remained adhering to them, hiding to some extent the brachial appendages. On examination of these fragmentary labial appendages with an enlargement of 20 to 30 diameters, one immediately observes calcareous spicula, the sharp extremities of which form a most elegant border, encroaching on the brachial membrane. These spicula were formed therefore from the beginning of the formation of the shell and, already very complicated, entirely resemble those seen in the adult individual . The cirri and the channel of the labial

spread out on the interbrachial

appendages are enveloped by a layer of spicula spread out