rent. The first and last dorsal spines are much shorter than the corresponding ones of carponemus, and the notch of the fin is conspicuous from the greater height of the soft rays. The second anal spine is very stout, and it rather exceeds the third one in length. The preorbitar is smaller than in that species, and its length does not exceed the diameter of the orbit. The face is therefore shorter, and the profile rises more steeply to the dorsal, owing to the greater height of the fish. The elongated pectoral ray, which is the tenth, reaches no farther back than the beginning of the anal. The scales are rather large and much tiled. About fifty-two exist on the lateral line, besides six or seven rows on the base of the caudal, and there are twentytwo rows in the height of the fish.

Mr. Lempriere, from whom we had the specimens, says that the fish is known at Port Arthur under the name of "the Perch," and has a bright silvery hue with dark spots. The specimens still exhibit many dark brown spots scattered thickly on the back and more sparingly on the sides, most of them being rather smaller than the exposed disk of a scale. The vertical fins, particularly the caudal, are more minutely spotted. The top of the gill-cover is blackish, and there is a dark mark on the humeral bone. As is usual in the genus, the inside of the mouth and lining of the gill-opening are purplishblack. Length 124 inches. Greatest height 3 inches.

The Cheilodactylus carmichaelis (Hist. des Poiss. v. 360) (Chatodon monodactylus, Carmichael, Linn. Trans. vol. xii. p. 500. pl. 24) approaches aspersus in shape, in the length of its long pectoral rays, and in the number of fin rays generally, but it is distinguished by six short, broad dark bars on the back. The formula of its rays is as follows:-Radü.-Br. 6; D. 17|24; A. 312; P. 9 et VI.; V. 15, Carmichael.

The Cheilodactylus fasciatus (Cuv. et Val. v. 357) of the Cape is distinguished by four or five vertical dark bands and five transverse lines on each lobe of the caudal. Its rays are stated to be:

:

Radii.—Br. 5; D. 19|23; A. 3|11; C. 17; P. 10 et V.; V. 15. Hist. des Poiss.

CHEILODACTYLUS GIBBOSUS, Solander. (Chatodon.)

Chaetodon gibbosus, Banks, Icon. Parkins. ined. t. 23. Cheilodactylus gibbosus, Richardson, Zool. Trans. vol. iii. p. 102. Radii.-D. 17 36; A. 38; C. 144; P. 8 et VI.; V. 15, spec.

(Pisces, Pl. II. fig. 3, 4.)

This fish inhabits the seas of Van Diemen's Land and the east coast of New Holland, as well as King George's Sound. A full description of it is contained in the Zoological Transactions' quoted above. It has the highest spinous dorsal of any described species of Cheilodactylus, and in the distribution of its black bands it bears a considerable resemblance to Eques americanus.

Mr. Neill gives a drawing of it (No. 24), and states that it is known to the aborigines of King George's Sound by the name of No. CCV.-PROCEEDINGS OF THE ZOOLOGICAL SOCIETY.

"Knelvek." The natives spear it on sandy banks, but say that it is rare. Its scales are smooth, and the second and third anal spines are moderately long and equal to each other. The suboperculum is narrow, and together with the other opercular bones and cheek is scaly. The figure is one-third of the size of the specimen. The scale is magnified. A considerable part of its disk retains the small asperities or ptenoid teeth, which do not however extend to the margin of the scale, that being, as is usual in the genus, thin and membranous. CHEILODACTYLUS NIGRIPES, Richardson.

Radii.—Br. 6; D. 18|26; A. 3|10; C. 138; P. 7 et V.; V. 15,

spec.

The aborigines of King George's Sound had no name for this species, and no drawing of it was made by Mr. Neill. The only specimen of it obtained was speared by a native named Murrianne, and measures 13 inches in length. It has a conical eminence on the prefrontal bone, like that existing in Ch. gibbosus; its face is short, with the profile ascending almost as much as in the species just named. The length of the preorbitar is rather less than the diameter of the orbit, the eye is placed midway between the gill-opening and mouth, and the interoperculum is only about half as wide as the disk of the preoperculum. The cheek and all the pieces of the gill-cover are densely scaly. The second of the simple pectoral rays is the longest and it falls short of the anus, while only about one-third of its length projects beyond the membrane. The spinous part of the dorsal is arched anteriorly. Its fifth and longest spine rather exceeds one-third of the height of the body. The preceding ones are graduated to the first, whose height is only a fifth part of the fifth one, but the decrease of the posterior spines is much less rapid, the last one having half the length of the fifth. The soft rays rise to nearly twice the height of the posterior spines, rendering the fin notched. The third anal spine is somewhat longer than the second one, which is stouter, but the spines generally are of moderate thickness, and are compressed. The caudal is forked to half its depth. The ventral spine is long and slender. The scales are without asperities, and the exposed part of their disk exhibits the concentric rings of structure distinctly. About sixty-one exist in a row between the gill-opening and caudal, exclusive of three or four on that fin. The teeth on the jaws are slender and closely set.

In the dried specimen the ventrals are pitch-black, and the other fins are nearly equally dark. The body is also dark, but in the absence of drawings or descriptions of the recent fish we cannot state its proper tints.

CHEILODACTYLUS ZONATUS, Cuv. et Val.

Cheilodactylus zonatus, Cuv. et Val. vol. v. p. 365; Rich. Rep. Brit. Assoc. 1845, p. 239.

Radi.-D. 1731; A. 38; C. 14§; P. 8 et VI. spec.

This fish, which is common to the China and Australian seas, appears to be called the "Zebra-fish" by the sealers who frequent

King George's Sound, though that name is most generally appropriated by them to the Crenidens zebra. Its prefrontal bone projects behind the nostril, but not so acutely as in Ch. nigripes or gibbosus. There is however a difference in this respect in different individuals. The width of the interoperculum is about half that of the preopercular disk, and these bones and the cheek are densely scaly. The scales of the cheek however are imbedded in spongy porous skin. The length of the preorbitar equals the diameter of the orbit. In the relative sizes of the opercular bones and preorbitar, and in the form of the dorsal, zonatus and nigripes closely resemble each other, but there is a difference in the anal spines, in the rays of the pectoral, in the shape of the caudals, that of zonatus being only sparingly excavated, and a striking one in the colours.

The dried specimen of zonatus shows very distinctly eight dark oblique bars on the body, the first crossing the nape and the last the base of the caudal, the intermediate pale spaces being equal to the bars in breadth. The entire head, including the preorbitar, is thickly marked by round dark spots of the size of duck shot. There are large spots on the caudal, which are so crowded on the margin of the fin as almost to form a continuous bar. Two or in some parts more rows traverse the dorsal, and there are dark marks on the tips of the anal and ventrals. The simple rays of the pectoral are orange. Mr. Reeves's drawing of the Chinese fish represents it as dressed in very lively colours during the breeding season.

The dorsal is highest at the fifth spine, as in zonatus, and is in other respects similar in form; but the anal spines are shorter, especially the second, which is also stouter in proportion. Rather less than one-third of the longest pectoral ray projects beyond the membrane, and the membrane is less deeply notched between the other simple rays than in nigripes. The scales differ from those of the lastnamed species, being finely granulated on the disk, as in nigricans.

The rays are somewhat differently enumerated in the 'Histoire des Poissons,' from a Japanese specimen. Radii.-Br. 6; D. 17|29; A. 38; P. 9 et V.; V. 15, Cuv. et Valenc.

The Cheilodactylus brachydactylus (Hist. des Poiss. p. 361) of the Cape approaches more nearly to our examples of zonatus in the numbers of the rays, but it does not appear to possess the prefrontal prominence, and has no other markings than a triangular black mark behind the eye. Radii.-Br. 5; D. 17|31; A. 39; C. 17; P. 8 et V.; V. 15, Cuv. et Valenc.

Cheilodactylus ciliaris, Richardson, Zool. of the Voy. of the Erebus and Terror, p. 37. pl. 26. fig. 6, 7 (Latris; Sciana ciliaris, Forster, &c.), is a species which is allied to the following ones, in the shortness of its simple pectoral rays.

CHEILODACTYLUS HECATEIUS, Richardson.

Latris hecateia, Richardson, Zool. Trans. p. 106. tab. 6. f. 1. Radii.—Br. 6; D. 1836; A. 327; C. 168; P. 9 et IX.; V. 1|5,

spec.

In the account of this species quoted above, I expressed doubts of the rank of Latris as a subdivision of the Cheilodactyli; but now that I have had an opportunity of examining a more complete gradation of specific forms, I am not disposed to think that it merits to be considered even a subgenus, though the non-prolongation of one of the pectoral rays (usually the tenth) makes it a convenient division of the Cheilodactyli, now known to be numerous.

This species inhabits the seas of Van Diemen's Land.

CHEILODACTYLUS LINEATUS, Forster (Sciana).

Cichla lineata, Schneider.

Sciaena lineata, T. R. Forsteri Descr. Anim. p. 134. An. 1844; Fig. pict. Georg. Forsteri in Bibl. Banks. servata.

Radii.—Br. 6; D. 18|36; A. 1|26; C. 30; P. 17; V. 15, Forst. l. c. This species agrees nearly with the preceding in the numbers of its fin rays, except that Forster says expressly that it has only one anal spine. It has also four dark dorsal stripes, with three intervening silvery ones; but it differs from hecateius in the yellowish colour of its fins, and particularly of its caudal, which obtained for it the appellation of "Yellow-tail" from the sailors. It frequents, like the other Cheilodactyli, rocky places, was captured by Cook's sailors with the hook, and was much approved as an article of food. It is a native of the seas washing the southern island of New Zealand. Length of specimen described by Forster, 24 inches.

Having seen no specimens we cannot institute a correct comparison with hecateius.

THREPTERIUS, Richardson.

(Openтhpios, ad alendum idoneus.)

Genus piscium acanthopterygiorum Cheilodactylis affine. Corpus catheto-plateum, ovato-oblongum, squamosum. Caput aliquantulum parvum, cute porosâ tectum, absque spinis, angulis vel aciebus serratis osseis. Os ut in Cheilodactylis extensibile. Dentes in premaxillaribus, mandibulâ trigonioque vomeris unâ serie instructi, brevissimi, parvi, subconici. Ossa palatis lævia. Genæ craniumque esquamosæ. Os preorbitale angustum. Operculum subtriangulare squamis tectum. Membrana branchiostega radiis sex curvis, satis validis sustentata. Squamæ læves nec dentatæ ; linea lateralis recta. Radii pinnarum pectoralium inferiores simplices. Pinna dorsi e nuchâ ferè usque ad caudæ pinnam regnans, squamulis apud radios instructa, membranâ inter spinas profundè emarginatâ ; lobulo tamen membranaceo e summis spinis pendente. Pinnæ ventrales thoracicæ sed a gulâ paulo remotæ.

The characters are deduced from dried specimens, and the pharyngeal teeth and structure of the intestinal canal are unknown. The jaw teeth are not strictly disposed in a single row, since a few minute ones form a row behind the others in front of the premaxillaries; but these can scarcely be visible in the recent fish. The chevron of

the vomer is acute and projects a little. The orifice of the mouth is rather larger than in the Cheilodactyli, but the jaws are extensible in about the same degree. The maxillary bone wants the flat thin plate near its head which exists in the Cheilodactyli and glides beneath the preorbitar. The latter bone is narrow, its width not being equal to one-third of the diameter of the orbit. The eye is comparatively large, three diameters and a half of the orbit being equal to the entire length of the head, and two of these diameters measure the distance between the hinder edge of the orbit and the tip of the gill-cover. The position of the eye is high enough to encroach upon the profile. The cheek equals the diameter of the orbit in breadth; the disk of the preoperculum is also wide, and the interoperculum moderately so. The operculum and suboperculum conjointly have a triangular form; the former is notched, and the latter is prolonged by a membranous tip, which forms the apex of the gill-cover. Both these bones are densely scaly; there is also a row of scales on the interoperculum, partially overlaid by the thin edge of the preoperculun., and the temples are also scaly. The rest of the head is without scales, but the mucous skin, full of canals and pores, which envelopes the head, prevents us from ascertaining the exact extent of the scales, at least in the dried specimens. The top of the head is destitute of scales to the occiput, but in the Cheilodactyli, dense, small scales extend forward on the skull to before the eyes. In the absence of thick fleshy lips, the genus differs from Cheilodactylus. The preorbitar is neither wide enough nor long enough to conceal the maxillary, which however enters partially beneath its edge. The thin crescentic border of the preoperculum is striated, but not crenated. The same kind of streaks or furrows may be discerned, though not so readily, in some Cheilodactyli. The head forms a fourth of the total length. The height of the body is also equal to a fourth of the length of the fish, caudal included. The belly is prominent, and the tail, posterior to the vertical fins, is slender. The lateral line is straight, and each of its scales is marked by a short straight tube, which is placed somewhat obliquely to the general direction of the line. About fifty-two scales compose a row between the gill-opening and caudal, the base of whose rays are also scaly, and the lateral line is prolonged as far as the scales extend on that fin.

The dorsal commences over the upper angle of the gill-opening and reaches to within an inch of the caudal. Its seventh spine, which is the tallest, is nearly equal to half the height of the body; the others are graduated very slightly posteriorly and more rapidly anteriorly. None of them are stout, and all of them are traversed on each side by a deep furrow. The membrane between them is deeply notched, as in the genus Pelors, and a slender process running up the back of each spine surmounts it in form of a small free lobe. The soft rays surpass the tallest spine a little, and are more than twice the height of the last one. The anal commences opposite to the beginning of the soft portion of the dorsal and ends beneath its tenth branched ray, or, in the specimens before us, about two inches and a half from the caudal. The spines are like the dorsal ones, grooved and slender,