The genus Myxine contains hagfish in which the branchial sacs on each side open to the exterior by a common branchial aperture which is ventral in position. Species of this genus are known from the Atlantic oceans, the Straits of Magellan, the coast of Chili, from Japan, and from South Africa. So far, there seems to be no record of the genus from Australian or New Zealand waters. During the clearing-away of a trawl on a fishing trip in Cook Strait in March, 1947, Mr. F. Abernethy noticed a small hagfish which, unlike the common blind-eel, "did not slime," and he preserved the specimen, which he later presented to this Department for examination. An account of this specimen was prepared as a thesis. The present description is based on a re-examination of the specimen, on data obtained from a second, smaller specimen which was found in the stomach contents of a "dogfish" taken off Kaikoura at a depth of 40 fathoms on 15th November, 1946. The latter specimen is partly damaged by digestion.

Bigelow and Schroeder (1948) have reviewed the genus Myxine]. They distinguish M. circifrons, M. garmani, and M. tridentiger as a group having the first three (the median) anterior lingual teeth on each side fused to a common base; and a second group having only the first two teeth so fused. The second group contains M. paucidens, M. affinis, M. glutinosa, M. capensis, and M. australis]. The Cook Strait specimens are to be included in this second group. M. paucidens has 6/7 lingual teeth on each side; M. affinis, 10 to 11 teeth in the anterior series; but M. glutinosa, M. capensis, M. australis have 7 to 9 teeth in each series on each side; these three species are not separable from the published accounts, although the vadidity at least of M. australis is certain, since J. R. Norman has handled material and maintains this species.

The difficulty of distinguishing the species allied to M. glutinosa arises from the ranges in the diagnostic data for M. glutinosa]. This species is known as having 6 and 7 pairs of branchial sacs; 7 to 9 teeth in the anterior Ungual series, 8 to 10 in the posterior; 53 to 70 abdominal mucous pores, etc. These ranges confuse the data for M. capensis (v. Barnard, 1925: 7 branchial pouches; 10 teeth in each series; 58 to 67 abdominal pores) and for M. australis (v. Norman, 1937: 6 pouches; 8 teeth in the anterior, 8 or 9 in the posterior; 56 to 64 abdominal pores). Our specimens have 7 branchial pouches; 8 lingual teeth in the anterior and 9 in the posterior series on either side; but are dearly distinct in the greater number of page 2 the mucous gland pores in the abdominal region and the high counts in the other regions. M. glutinosa has 26 to 33 pores anterior to the branchial apertures, 53 to 70 in the abdominal region, and 11 to 13 posterior to the vent—a range, all told, of 80 to 116. The range for M. austral is falls within those figures, there being 30 to 36 + 56 to 64 + 9 to 12, a range of 95 to 112 for this species. Likewise, in the case of M. capensis, where the mucous gland pores are arranged as 28 to 31 + 58 to 67 + 10 to 13—a total ranging from 96 to 111.

The larger of our specimens, which is only 315 mm. in total length, has 44 + 95 + 25 actual glands. The number of pores is difficult to count; but this total of 164 glands is only one more than the number of pores counted, and greatly exceeds the total in any other known species. The count is lower for the smaller specimen, which is, at the best estimate, 163.0 mm. in length, coiled and with the skin digested and damaged in the posterior third of the abdomen and over part of the tail. Combining the count of pores, where the skin is in good condition, with that of glands where these are exposed gave 36 + 84 + 17—a total of 137, which is still above that of any other species. In both specimens, the mucous glands are not spaced out from one another, but crowded so that they present

Text-figures

1—Ventral view of the head and branchial region showing the ventral fin (V.F.), the lateral fins (L.F.), and the branchial apertures (B.A.). 2—Ventral view of head showing the nasal (N.), subnasal (S.N.), lateral (L.), barbels and labial tentacles (L.T.). 3—The array of lingual teeth.

flattened anterior and posterior faces rather than being rounded in outline, as illustrated in other species. Also, in both specimens the glands are not segmental in number or arrangement. For example, in a space of ten myotomes in the abdominal region of the larger specimen, there are 15 glands, and in the smaller specimen 13 were counted in the equivalent length.

External features have not been found satisfactory in separating one species of Myxine from another; but both of the present specimens show a unique external feature in the form of an anterior continuation of the ventral (precloacal) fin which extends between the branchial apertures as a fin diminishing in height to become a low ridge or fold which is quite definite and does not terminate until about half-way to the lower lip. Equally distinctive is the presence on either side of a low but well-formed fin which extends from close to the ventral fin immediately posterior to the branchial aperture in an oblique anterodorsal direction on to the side of the pharyngeal region to terminate just ventral to the row of mucous gland pores at a level nearly two-thirds to the anterior end of the ventral fin. Being truly pharyngeal in position, these resemble metapleural folds.

This occurrence of an anterior extension of the ventral fin, the presence of the lateral fins, and the non-segmental arrangement of the mucous glands are contrary to the generic definition; but no more so than some features in other species. Moreover, since these specimens show agreement in general facies and anatomy, especially of the pharyngeal region, we have no hesitation in referring this species to the g. Myxine as a new species named M. biniplicata on the basis of the lateral pharyngeal fins.