7. Criticisms on some points in the Summary of the Results of the 'Challenger' Expedition. By Dr. H. O. FORBES.

8. Observations on the Marine Fauna of Houtman's Abrolhos Islands, Western Australia. By W. SAVILLE-KENT, F.L.S., F.Z.S.

Mr. Saville-Kent's investigations of the marine fauna of Houtman's Abrolhos Islands were associated with a visit he paid them in his capacity of Commissioner of Fisheries to the Western Australian Government, and were conducted with the particular object of advising that Government as to the conditions and prospects the adjacent waters presented for the establishment therein of profitable oyster or mother-of-pearl shell fisheries.

Houtman's Rocks, or Houtman's Abrolhos, as they are variously charted, are so named after one of the early Dutch explorers in contradistinction to a coral group, also known as the Abrolhos, lying off the coast of Brazil. The island group discussed in this paper is a small archipelago, chiefly of coral origin, situated between the latitudes of 29° and 30° S., about thirty miles west of Champion Bay and the important Western Australian port of Geraldton.

As a result of his investigations Mr. Saville-Kent found that the ordinary Australian rock oyster, Ostræa glomerata, occurred there in tolerable abundance and under conditions that would justify its being made the subject of systematic cultivation. The smaller West Australian variety mother-of-pearl shell allied to or identical with Meleagrina imbricata occurs very sparingly on the Abrolhos Reefs, but in the Commissioner of Fisheries' opinion was not worthy of serious attention in face of the unexpectedly favourable conditions he discovered to obtain there for the introduction and acclimatisation of the larger and more valuable species, Meleagrina margaritifera. This decision was arrived at as the outcome of an investigation of the associated marine fauna, and which was found to present features of high interest from both a utilitarian and a biological standpoint.

The existing pearl and mother-of-pearl shell fisheries of Western Australia, as associated with the larger species, have not hitherto extended further south than Exmouth Gulf, in about lat. 22° S., and are consequently limited to the Tropics. The fishery for the smaller species, Meleagrina imbricata, is confined chiefly to Shark's Bay, three to four degrees south of Exmouth Gulf, and has in consequence of the wasteful depletion of the banks in former years been reduced to a comparatively low state of productiveness. Among other operations initiated by Mr. Saville-Kent, with the object of resuscitating the Shark's Bay fishery, has been the experimental transportation to it and cultivation of the large tropical pearl shell Meleagrina margaritifera. These acclimatisation experiments, although initiated only on a small scale, have been attended with complete success. The large mother-of-pearl shell has not only shown its capability of thriving in the colder waters of Shark's Bay, but has within a year of its transportation to this extratropical area commenced to freely propagate.

The site selected for the foregoing experiments in Shark's Bay was the neighbourhood of extensive banks of coral growths pertaining to the genus Turbinaria. and from which reefs Mr. Saville-Kent obtained the remarkably large specimens of this Madrepore that are now on view in the exhibition galleries of the Natural History Museum, South Kensington. It has been determined by Mr. Saville-Kent in the course of his Australian explorations that the genus Turbinaria represents the group of Madrepores which in Australian waters enters most extensively into the composition of coral reefs in the colder or extra-tropical limit of their distribution. This predominance of Turbinarians had been found by him to obtain at Wide Bay, on the southern outskirts of the Great Barrier Reef in Queensland; in the colder though more northern waters at the head of the Gulf of Carpentaria; and, finally, in the Shark's Bay district of Western Australia.

The conditions which permitted the successful acclimatisation of Meleagrina margaritifera in Shark's Bay were found by Mr. Saville-Kent to be still more favourably fulfilled around Houtman's Abrolhos. In and among this island group,

notwithstanding the fact that it lay some two degrees south of Shark's Bay, the character and composition of the coral reefs proved to be entirely distinct. In place of the extra-tropical Turbinaria the corals of the Abrolhos Reefs comprise, as in essentially tropical districts, numerous varieties of branching Madreporæ, or so-called Stag's-horn corals, commingled with many species of Porites, Montipora, Pocillopora, Seriatopora, Coloria, Mussa, and other intra-tropical reef-building species.

A yet more remarkable phenomenon, however, is recorded by Mr. Saville-Kent in connection with the marine fauna of Houtman's Abrolhos. This is the circumstance that he discovered on its reefs three of the most valuable economic species of Holothuridæ or Bêche-de-mer, identical with types that are systematically collected in Torres Straits, and throughout the northern moiety of the Queensland Great Barrier Reef, but which are unknown to the coastal reefs of Western Australia further north, and where their place is taken by a distinct and much less valuable commercial species.

The fish fauna of Houtman's Abrolhos, while corresponding to a large extent with that of the temperate Australian seaboard, as instanced by such genera as Pagrus, Aulopus, and Seriola, is also associated with many essentially tropical species, including, notably, a large assemblage of brilliantly coloured Labridae, or Parrot-fishes. Certain of these Labridae, while not obtained by Mr. Saville-Kent in collections made among the mainland reefs of Western Australia, were familiar to him, as in the case of the Holothuridæ, as denizens of Torres Straits and the northern region of the Great Barrier Reef.

ocean

The anomalous character of the marine fauna of Houtman's Abrolhos as herein defined can only be accounted for by the assumption that an current setting in from the equatorial area of the Indian Ocean penetrates as far south as this island group, and has borne with it the floating embryos of the Holothuridæ and Coelenterates, &c., that so characteristically distinguish it. A reference to the Admiralty charts, dealing with the ocean currents of this region, supports this interpretation to a considerable extent; indicating, as a matter of fact, a prevailing northerly set along the western coast of Australia, but at the same time a distinct southerly intrusion of the waters of the Indian Ocean at some distance off shore down towards and closely approaching Houtman's Abrolhos. In further support of this interpretation Mr. Saville-Kent also determined by synchronous readings of the thermometer at the coldest season of the year, July, that as great a difference as from ten to fourteen degrees Fahrenheit distinguished the surface temperature of the sea at respectively the Abrolhos Islands and in Champion Bay. Mr. Saville-Kent remarks, in conclusion, that much scope is yet left for further investigation in this direction; while with respect to the anomalous character of the marine fauna it would be greatly to the advantage of marine biological science if a thoroughly exhaustive investigation thereof could be carried out.

9. On Hereditary Polydactylism. By Dr. GREGG WILSON.

10. On the Reproduction of the Common Crab. By Dr. GREGG WILSON.

TUESDAY, SEPTEMBER 17.

The following Papers were read:

1. Observations on Instinct in Young Birds.

By Professor LLOYD MORGAN, F.G.S., Assoc.R.S.M.

This paper dealt with observations on young moorhens, chicks, martins, and swallows with a view to determine how far the activities involved in locomotion

(swimming, diving, running, flying), in feeding, bathing, &c., are instinctive or congenital in their definiteness; and how far the definiteness of these and other activities is a matter of individual acquisition. Observations were also made on congenital and acquired timidity. While the performance of these activities has a congenital basis they are perfected by individual acquisition. There is no instinetive and congenital avoidance of insects with warning colours; that appears to be entirely the result of individual experience. There seems to be little or nothing in the observations to afford any material support to the view that the instinctive activities result from the inheritance of what is individually acquired.

2. Notes on the Early Development of the Ganoids, Lepidosteus, Acipenser, and Amia. By BASHFORD DEAN, Instructor in Biology, Columbia College, New York.

A. Segmentation of the Egg.--The earlier cleavages conform to the usual plan of Teleost and Amphibian :-Lepidosteus and Amia meroblastic, Acipenser superficially holoblastic. Questions as to the kinships with the yolk type of the Elasmobranch on the one hand, and with that of the Teleost on the other, were discussed.

B. Blastula, Gastrula.-The relations of the different forms of Ganoidean blastula were shown in diagrams. The blastula of Lepidosteus and Shark, of Amia and Teleost are similar. Comparison of Ganoidean gastrula: the diagrams show structures diverging from the type of Lepidosteus towards that of Teleost.

C. General Mode of the Formation of the Embryo.-Shark-like characters of Lepidosteus, flattened growth of Acipenser, and Teleostean features of Amia.

D. Conclusions.-Developmental nearnesses of Lepidosteus to the Elasmobranch and of Amia to the Teleost, and the evidence on the side of embryology for connecting the line of the Teleosts with that of the Ganoids, as well as for drawing more closely together the Elasmobranchian and Ganoidean phyla.

3. On some questions relating to the Morphology and Distribution
of Medusa. By Dr. OTTO MAAS.

Dr. Otto Maas exhibited some plates from his monograph of the 'Albatross' Medusa, and discussed some questions arising from the study of these Pacific forms. The collection, though not very rich, is of interest in various points:

1. Morphological.

2. Zoogeographical.

3. Bionomical.

1. Amongst 18 species 9 are new, several of them peculiar forms, for instance, a representative of the aberrant genus Homoioneme, established 1892 for some forms of the Plankton' Expedition. Amongst the Acraspeda we find the genera Periphylla, Atolla, and others which are of importance for the morphology of the whole group, and which have induced Claus and Vanhöffen to a reformation of Häckel's system. The previous authors could not study the genital and sense organs; a detailed study of these shows that we can trace a line of relationship from the primitive Lucernarida through forms like Periphylla and Nauphanta to the higher Discophora, forms like Atolla lying a little to the side of the line, whilst Charybdea is totally away from it. The study of the canal system of the Periphyllidæ and their relations shows some primitive features in correspondence with the embryology of the higher forms, i.e., the interruption of the continuous entodermic cavity at four interradial points by the invagination of the Trichterhöhlen.'

2. The Medusæ have been caught in an oceanic basin hitherto scarcely explored. In a map of the distribution of the Cathammata given by Vanhöffen the part of the Pacific navigated by the Albatross' is an empty gap which is now filled up.

The list of Acraspeda species shows a striking resemblance to that of the 'Challenger' Expedition. The so-called 'deep sea Medusa' seem to have a very

6

wide geographical distribution; they have been brought home by the 'Challenger,' the Vettor Pisani,' the National,' and the Albatross,' but it is to be noticed that they have been caught only in those really oceanic explorations, so, if they are, not deep-sea Medusa, they are certainly not forms of the shallow water.

Amongst the Polypomedusa we find a very close relationship between Atlantic and Pacific species. The different species of one genus are in general much more difficult to distinguish than amongst the Trachomedusæ. This perhaps may be explained by the effective power of passive dispersal.

3. For the first time a great number of sketches of living material of the Periphyllidæ, &c., had been obtained on board. All these show the dark purple colour, generally attributed to deep sea animals. The explanation for other forms is, that in the green phosphorescent light of the abysses, purple is the complementary colour, which makes the animal invisible, and acts as a protective colour. It would be dangerous to conclude from this that Periphylla, &c., are deep-sea forms. They have been brought up in an open trawl from a great depth, but the closed part of the net contained no Medusa. If a haul from a great depth contains forms which did not occur in surface hauls, these forms do not necessarily come from the abyss, for they might have been caught on the way to the surface. Our knowledge of the pelagic life of the surface is still so incomplete that every expedition brings us new species, as has been shown in the Copepoda, Medusa, and other groups of the 'Albatross' Expedition.

4. On the Spermatogenesis in Birds. By J. E. S. MOORE.

The observations were made to ascertain whether the course of the spermatogenesis in birds was essentially similar to that of other vertebrates recently examined. It was found that in two points of chief importance, namely, the manner of numerical reduction of the chromosomes and the alternation of the homo- and heterotype divisions, the spermatogenesis of birds is closely similar to that of the remaining vertebrate forms.

During the first heterotype division, which corresponds to the division of the growing cells in Mammals and of the great spermatocytes in Elasmobranchs and Amphibia, the spermatic elements of pigeons show a marked tendency towards the formation of multinucleate masses. One of the most interesting features appertaining to these bodies is that the spindle-figure during the division of their nuclei appears to originate entirely within the nucleus, since the nuclear wall can be distinctly seen after the spindle-figure has been fully formed. The stages in the division may be diagrammatically represented thus:

The advent of the great heterotype mitosis is always preceded during the spermatogenesis by the peculiar convoluted and lop-sided figure (a which is

here, as elsewhere, characteristic of what I have previously termed the 'synaptic phase.'

The whole course of the spermatogenesis appears to correspond more closely with that of Elasmobranchs than of Mammals, since there appear to be two generations and one division after the synapsis before the spermatozoa are complete.

As in Elasmobranchs and Mammals, the number of the chromosomes appears to be reduced during the synapsis, and to be then determined for succeeding divisions, just as in the case of plants.

The spermatogenesis of birds supports in every way the conclusion first put forward by Strassburger, which is at present gaining ground, namely, that the process of numerical reduction in the chromosomes is not brought about by any division at all, and is similar for both animals and plants.

5. On the Development of the Teeth in Certain Insectivora.
By M. F. WOODWARD, Demonstrator of Zoology, R.C.Sci. Lond.

In the hedgehog the author describes vestigial calcified milk predecessors to the third upper incisor, the lower canine, and the first pre-molar of both upper and lower jaws, and an uncalcified vestige of the milk predecessor of the second lower incisor, thus extending Leche's observations and confirming his later conclusion that the adult incisors, canines, and pre-molars all belong to the third or replacing tooth series. In addition, a vestigial anterior lower incisor and a third lower pre-molar were observed. Indications of three dentitions are described for the molar series, the molars being referred to the third or replacing dentition.

The teeth of Gymnura, Sorex, Talpa, Centetes, and Ericulus are also dealt with, and the following points more especially noted:

1. The presence in Gymnura of five pre-molars in both upper and lower jaws, represented in both dentitions.

2. The absence of the alleged milk predecessor to the first pre-molar of Talpa described by Spence Bate, that tooth being shown to be itself a milk tooth.

3. The development, in all cases, of the successor to the fourth pre-molar between the 'deciduous pre-molars 3 and 4.' The facts associated with this appear to indicate that the so-called 'deciduous pre-molar 4' is a precociously developed molar, and that the tooth which replaces it is a much retarded premolar of the milk series.

Two sets of calcified teeth are shown to be for the greater part developed among Insectivores, and it is characteristic of them that there is a tendency towards reduction of the milk set with early development of the replacing dentition.

6. On the Mammalian Hyoid.1 By Professor G. B. Howes.

The author proved from the study of Nasua that the small bone attached to the paroccipital process in Lepus and Procavia (Hyrax), independently described by Krause and Brandt, is in reality the styloid, and showed that the discovery enables us to recognise two distinct culminating types of modification of the hyoid of mammals, viz. (i.), the protero-stylic, known only in man and the marmosets, and (ii) the opistho-stylic, known only in the rodents mentioned. Reviewing the subject more generally, he called attention to the presence of a considerable tympano-hyal, occupying a novel position, in Cholapus, and he exhibited the hyoid of a young rabbit, the body of which was subdivided by a transverse suture, probably indicative of the original demarcation-line between its two component copulae. A classification of the types of mammalian hyoid was submitted.

1 Paper will be published in Jour. Anat. and Phys., Jan., 1896.