at posterior end, which reaches slightly beyond vertical line drawn from anterior margin of orbit. Mandible projects a little beyond intermaxillary. Mouth, when fully open, has a gape of three-quarters to an inch. Opercula all well defined, posterior margin nearly vertical and straight, five or six distinct striæ on preoperculum pointing to base of pectoral fin. Diameter of eye one-fifth length of head, pupil black, iris olive-yellow. This organ—the eye—is sunk with orbit about one-eighth of an inch (or below the plane of the cheek), and is protected by a beautifully-transparent disc of apparently thickened skin, convex in form, extending half an inch in front of centre of eye and three-tenths of an inch behind the same point, and provided with a slit or opening directly over the centre of eye vertical in position and gibbous in form or spindle-shaped. This disc is evidently meant to protect the eye, and, at the same time, to compensate for the diminution in the range of vision caused by the eye being sunk in the head. Fins small and delicate in structure, rays mostly soft or branched. One dorsal fin only, situated exactly midway between snout and origin of middle or shortest rays in tail-fin, fourth ray longest. Ventral fin origin, in vertical line from eighth or middle ray of dorsal fin; tail-fin deeply forked, nearly bifurcate. The posterior end of the anal fin is distinguished by a few long feathered rays much longer than those adjoining them in body of fin. The scales are large, irregularly rounded, transparent, and on exterior margin pectinate, also very tough, and not deciduous. Along the base of or parallel to base of the dorsal, pectoral, and ventral fins, rows of scales are situated differing in form from the body-scales of the fish, but resembling in shape those fins beside which they grow; and on each lobe of the caudal fin on both sides are scales or skin-like plates, somewhat like the respective lobes they are attached to. The other or body scales, besides possessing the shape described above, are arranged in layers which overlap each other so as to leave a diamond pattern over the surface of the trunk of the fish. As in the pilchard of the English channel, described by Yarrell, a series of three-limbed scales, or rather dermo-hæmal plates or processes (for they are bony) exist along the outline of abdomen from the pectoral to the anal fin. These are as figured, are buried underneath the true scales, they diminish in size towards the tail, overlap each other, and are placed with short end of mid portion pointing towards the tail. Looked at transversely they conform to the precise outline of that part of the hæmal arch here situated, viz., the bottom. No lateral line visible on this specimen.

In colour, indigo blue on back and head, shading off into grey on sides, and silvery white on belly, general aspect of body very silvery. There are seven or

Teeth.—None perceptible to touch, but mandible rough.

Contents of stomach.—Mass of brown granules same as in No. 2 and 3. Air-bladder silvery. Pyl. cæc., 69, but so much decayed as to be doubtful.

In April last, a fish caught at Otago Heads was given by Mr. Jewit, fishmonger, Dunedin, to Mr. Hugh Maclean, who, on having it cooked and eaten, was of opinion that it was a real herring. I did not see it, and it is impossible now to decide its species, only that I understand the fishmonger called it a Picton herring, which it probably was. However, on the 20th November, 1882, I got possession of an undoubted Picton herring from Mr. Maclean, and which had been taken in Otago Harbour; and the following is a description of it:—

[The section below cannot be correctly rendered as it contains complex formatting. See the image of the page for a more accurate rendering.]

In form, fusiform, dorsal outline nearly straight, belly curved from mouth to ventral fin deeply; head triangular in profile and transversely, interorbital space straight, broad and flat; fine striæ on operculum pointing to origin of pectoral fin, opercula same as in above specimens; mouth small and opening to three-quarters of an inch; maxillary fine, curved and with circular end; mandible projects slightly; eye 6/20 inch in diameter and situated close to ridge or interorbital space, pupil blue, iris yellow, depressed in cavity of orbit, transparent covering disc same as in specimen No. 1., only the opening or slit was much wider being equal to half diameter of eye. Fins fine, clear and delicately formed, dorsal and pectoral finely pointed; origin of ventral at vertical from base of eighth dorsal ray; caudal or tail-fin very much forked and having four scales as in above examples. No serrature of abdominal outline, but bony scales present or dermo-hæmal processes under true scales as figured above, overlapping and having short end pointing tailwise from pectoral to anal fin. No lateral line visible, but eight distinct bands parallel to axis of fish from gill openings to base of caudal, blue in colour along back and more or less distinctly spotted with marks of an indigo hue; those towards the belly lighter and greenish in colour, but without spots. In colour, back an indigo, sides and belly white, but the whole with a covering of silvery scales, nacreous under certain lights.

[The section below cannot be correctly rendered as it contains complex formatting. See the image of the page for a more accurate rendering.]

Dimensions.—W. 2½ oz.; L. 8 1/10 in.; D. 1 4/10; G. 3 6/10; l. d. of T., 5/10; head, 1 6/10.

[The section below cannot be correctly rendered as it contains complex formatting. See the image of the page for a more accurate rendering.]

Fin rays.—D, 18 or 16²; P, 17; V, 8; A, 17; C, 19; Br., 7. Lengths—D, 1 1/10; P, 1; V, 6/10; A, 9/10; C, long ray, 1 3/10, s.r., 3/10, spread, 1 9/10.

Teeth.—None perceptible to touch about the mouth anywhere.

Scales.—Lat. 1. 60; trans. 1. 12; large and irregularly rounded, overlapping.

Vert.—36, but incomplete owing to some getting astray before examination, and so lost.

Contents of Stomach.—Walls of abdominal cavity black in colour. Stomach siphonal, and containing mass of brown granules. Intestine full of olive-coloured softer matter. Pyl. cæc., 59; but, as these are small and delicate, and I used considerable pressure with back of knife in removing fatty matters to facilitate counting, I may possibly have removed a number without being aware of it. When cooked this pilchard was most excellent to eat. In several smoked specimens of this fish as sold in the Dunedin market I found, so far as possible to make them out, all the marks correspond with those of fresh specimens above described. Form of the head and gill-covers, striæ, position of ventral fin, fin-rays, all agree—the vertebræ in two examples numbering 50 each; but outline of abdomen was distinctly serrated or marked by raised scales, due probably to projection of dermo-hæmal plates after curing.

A comparison of our New Zealand pilchard or Picton herring with Yarrell's account of the English form shows such a close relationship as almost amounts to identity of species. It is also very interesting to notice how well designed certain parts are to fulfil their special functions, as the transparent jelly-like disc or covering for the eye. The eye being well sunk in the orbit beneath the plane of the cheek its range of vision would be very limited were the orbit not likewise sunk. This being so it is also necessary that the surrounding orbital bones should be gradually curved in to the depressed eye. This secures range, provision for which over the anteorbital bone is greater than over post-orbital, showing that the fish needs to see more ahead than behind. Then covering the eye is the disc I have mentioned, protecting the eye from injury, while it permits free vision by its transparency, with direct vision in front of fish by refraction, and by its form and bulk giving symmetry and completeness to the adjoining parts. From the difference in width of openings or slits between the first set of fish examined and the last one, I should expect that these fish have the power of opening and closing the slit at pleasure.

I have stated that the scales are tough and non-deciduous, and may add that they are so wonderfully overlapped and wedged together as to form an outer covering or coat of mail completely surrounding the trunk of the fish. This protection is a very obvious part of the design, for the bones of the skeleton are extremely fine and seem unequal (unassisted) to carrying the fleshy parts of the body. The abdomen in particular is a most delicate part, and was more or less injured in the fresh specimens examined by me, a characteristic which I found extending to the viscera also, to the prevention of my searches in that direction to some extent. And here again the perfection of design appears, for along the abdominal outline where the