The Plankton and Benthos animals are (with some exceptions discussed) more stenothermal; hence results the known division of the coasts.

A distribution of the animals of the open ocean into regions is possible too, chiefly on account of the currents.

Some instances of that were given from the facts of the Plankton expedition, chiefly of the Medus, which have been worked out by Dr. Maas. Most interesting in this regard are the Geryonids, which in every ocean basin do not exceed a certain N. or S. latitude, and which as Plankton animals occupy a similar portion of the map as the corals of the Benthos.

Dr. Maas is led to the conclusion that a comparison between the vertical differences and those found in higher latitudes can only be carried to a limited depth.

The horizontal distribution of the pelagic fauna is not compensated by the vertical differences.

There are no various belts of vertical life; the intermediate fauna between the Plankton of the surface and the Benthos of the abyss may be supposed to be only Nekton.

This is valid for the deep sea. In lower seas the life may be continuous from the surface to the bottom. Sometimes special conditions, especially of temperature, prevail (e.g., in the Mediterranean, where a depth of 2,000 m. shows 13°). The occurrence of a deep-sea pelagic fauna, neither belonging to the surface nor in connection with the ground, as well as the survival of deep-sea animals coming to the surface, may be explained by this higher temperature.

5. Second Report on the Zoology of the Irish Sea.
See Reports, p. 318.

6. On Marine Fish-hatching and the Dunbar Establishment of the Fishery Board for Scotland. By Professor W. C. MCINTOSH, F.R.S.

It was stated that we are yet in doubt as to the beneficial effects to the fisheries of the artificial hatching of sea-fishes, but that the importance of the issue demanded a thorough trial. Several nations, such as the Americans and Norwegians, had chiefly experimented with the cod, other forms having been dealt with in small numbers. Though the sole was selected as the most suitable species, the lateness of its spawning period gave an opportunity for a preliminary series of experiments with the plaice. Accordingly a total of 396 plaice were collected, the average size of the males being about 17 inches, and that of the females about 20 inches. From these 27,350,000 ova were obtained, hatched in the boxes (Dannevig's), and the larval plaice-to the number of 26,060,000-sent into the sea, the loss in the process being only about 4.4 per cent.

At Dunbar the eggs are shed in the spawning-pond, and carried to the spawncollector by the current. They are then counted and placed in the hatchingboxes. The sea-water for these is passed through a series of flannel-filters, so as to secure purity. Moreover, besides the current entering the end of the box, an upand-down movement is communicated to them twice every minute, so that the ova are evenly distributed through the water. On the whole the operations for the first year were most successful.1

1. On the Correlation between Root and Shoot. By Professor L. KNY.

A full account of the experiments will be given by Dr. Fulton in the forthcoming Twelfth Annual Report of the Fishery Board for Scotland.

2. On the Sensitiveness of the Root-tip. By Professor W. PFEFFER.

3. Exhibition of Diagrams. By Professor L. KNY.

1. Interim Report on Telegony.-See Reports, p. 346.

2. On some Difficulties of Darwinism.

By Professor D'ARCY THOMPSON.

3. On Social Insects and Evolution. By Professor C. V. RILEY. Experiment and discussion on the question as to whether acquired characters are transmitted or not through heredity have of late been largely based upon the economies of insects, and especially of the social species. The author gives a summary of what is known of the habits and economies of bees, wasps, ants, and termites, especially as to the development of the young. He points out that the origin of neuters, with their diversified forms, in these social insects has been considered one of the greatest difficulties with which the theory of natural selection has had to contend. Weismann, in urging his own particular theories to account for the variation which organisms have undergone, insists, and has, within the last year, in his controversy with Herbert Spencer, emphasised his belief, that these neuter insects absolutely preclude the idea of the transmission of acquired characters. The author believes, on the contrary, and endeavours to show, that while these neuters among social insects, with their varied structures and habits, do indeed offer serious obstacles to the theory of natural selection as an all-sufficient theory toexplain the phenomena, these are nevertheless perfectly explicable upon the general principles that have governed the modification of organisms, among which natural selection plays an important but limited part.

Among the social Hymenoptera, where, as in the bees and wasps, the larva is nursed and brought up in a definite cell or cradle, the three castes of male (or drone), fertile female (or queen), and neuter (or worker) are quite definitely fixed and separated. The differences between the worker and the queen are, however, solely due to the treatment of the larvæ, and are consequently under the control of the colony. The same larva, according to treatment and nurture, may produce either a perfect queen or a worker, between which the differences as to size, structure, external and internal organisation, and length of life are very great. This is absolutely and definitely proved for the bees, and is doubtless equally true, though with less absolute proof, of the wasps, in which the same three castes of male, female, and neuter obtain in some species, while in others the neuters are replaced by parthogenetic or unimpregnated females, normally capable of reproducing. In the ants, where the larva is not confined to a definite cradle, and where there are, in the more typical species, two castes of neuters, viz., soldiers and workers, the variation between the different castes is greater, and there is also more variation in the individuals composing the different castes; but the evidence all points to the fact that these different individuals are also the result of food and nurture, very much as with the bees and wasps.

In the three families of social Hymenoptera above mentioned the young are maggot-like and absolutely helpless and dependent on the nurses. In the termites, 1894.

YY

which belong to a different order (Platyptera), much older in time according to the paleontological record-an order in which the young are born in the image of the parent and are more or less independent from birth-one would expect to find larval nurture and environment less potent in influencing ultimate structure. Yet all the facts known, and particularly the late most painstaking observations and experiments of Grassi, prove conclusively that here, also, the young are dependent upon the nurses, and, more remarkable still, may be diverted, according to the food and treatment given, to any of the four castes which characterise the typical termite colony, there being, in addition to the male and female, two kinds of neuters, viz., soldiers and workers, as in the true ants. In the first larval stage, or when first hatched, the individuals are, to all appearances, absolutely alike, and each possesses the potentiality of becoming either a worker, or a soldier, or a perfect sexed individual. Nay, further, the pupa, or nymphs, may be diverted into reproductive forms which never acquire wings and which are called supplementary queens and kings; and even larvæ may be so diverted into reproductive forms, with no further external structural development, when they become complementary or neotinic kings and queens.

The steps in the development from the simpler to the more special structures and attributes belonging to the species with the most perfect social organisation may be traced in the different species and genera of their respective families in all social insects of the present day. The amount of variation is often great in the ants and termites, where the environment is less fixed than in the bees and wasps, and this variation, among termites, is particularly manifest in the economy of the same species, as exemplified in Eutermes, which the author has studied in the West Indies, and in which the number of queens varies from one to nine or more. It is not generally known, but it is a fact, that existing termites (using the term in the broader sense, so as to include several genera) exemplify all the steps in development from species which are active in broad daylight (the neuters having faceted eyes and dark integument, and, so far as is known, no definite nest or termitary), to the more specialised species in which the economy and division of labour are most perfect, and in which the neuters and soldiers are blind and always work in the dark and build elaborate structures. Further, the neuters in termites are truly without sex or are modified individuals which might have produced either sex; while in the Hymenoptera they are invariably modified females.

In so far as these different forms of neuter insects depend for their development on the food and treatment given by the nurses, they are outside the domain of natural selection. The author believes, however, that there is a potential, inherited tendency in the young larva to develop in the various directions that have been fixed for each species in its past development, as he cannot believe, e.g., that young larvæ taken from one species of termites, and brought up under the care of the nurses of any other species, can be diverted to the forms peculiar to this last. There is a possibility, since the food of these young in the social insects consists largely of secretions from the nurses, that these secretions may so influence the changes as to confine them to the specific forms of its own species, regardless of the parentage of the young. That there can be any such powerful influence of nurture as would neutralise and overcome the inherited tendencies of species is, however, extremely improbable: its bare possibility opens up a most interesting field for experiment, which is easily made, and doubtless soon will be made.

The author believes, with Darwin, that the variations in social insects have been guided by natural selection among colonies, but that there has also been what he calls social selection among individuals. Competition has been between colonies rather than individuals, and those colonies which have acquired, through heredity, the habit of producing, from one or more fertile females, the different castes characteristic of the species have, in course of time, survived. He believes, however, that this colony-selection, as well as the social selection among individuals, has been not only along lines that were and are useful to the species, but along lines of secondary utility, and even along lines which are purely fortuitous and still most variable and unfixed.

Finally, as between Weismann's views and those held by Darwin himself, the

author feels that the facts furnished by the social insects strongly favour the transmission, through heredity, of acquired characters, both psychic and .structural, but that they also require other factors besides that of natural selection to explain them.

The trouble with all theories of reproduction and heredity based solely on embryologic and microscopic methods is, that the essence, the life principle, the potential factors, must always escape such methods. Any theory that will hold must cover the psychical as well as the physical facts-the total of well-established experience and this truth was recognised by Darwin in framing his tentative theory of pangenesis. We are all in these matters simply discussing processes, and the author believes that too much has been made of the cell theory, the cell being but the medium through which assimilation, growth, organisation, regeneration, and reproduction are effected by the ultimate elements and the inherited potential forces, call them what we may. The idea that the individual during its lifetime develops all that is potential in the germ seems to him more philosophic than the idea that the germ originates, at a specific moment of time, the tendency to all that develops in the individual. It may be a perfectly correct conception, to use Weismann's language, that the primary constituents for the characters of the different forms of social insects are included in the egg, and that a particular form of stimulus decides as to which group shall undergo development; but it is difficult to believe, in the light of the facts which are known concerning social insects, that the different kinds of ids and determinants which are thus conceived to characterise the germ have not been impressed upon it as a consequence of the characters, both acquired and congenital, of the parents.

The author finally calls attention to the significant fact that just as in man, among Mammalia, the higher intellectual development and social organisation are found correlated with the longest period of dependent infancy; that this helpless infancy has been, in fact, a prime influence in the development, through family, clan, tribe, and state, of our highest organisation and civilisation; so in the insect world we find the same correlation between the highest intelligence and dependent infancy, and are justified in concluding that the latter is, in the social insects as in man, in the same way a prime cause of the high organisation and division of labour so characteristic of them.

4. On the Rôle of Sex in Evolution. By JOHN BERRY HAYCRAFT, M.D., Professor of Physiology, University College, Cardiff.

While Weismann admits that the power to vary is a property of protoplasm, he looks upon sexual conjugation as a means whereby the number of these variations may be increased. But we find that the power to vary is itself a most varying character of protoplasm, for both in the case of sexual and asexual reproduction certain species constantly produce striking varieties, while others rarely produce them. We can therefore state, not only that variation is a quality of protoplasm, but that it has and can acquire this quality in varying degree and apart from sexual congregation; moreover, greater varieties in the progeny can be obtained simply by increasing the number of the offspring. It would appear therefore that, if we accept the view held by Weismann, we must assign to sexual conjugation a function already possessed by protoplasm, and it is difficult to understand its utility. If, however, we remember that the varieties which occur in newly allied forms (the only ones which conjugate) are variations chiefly in quantity (differences in size of the whole or parts, amount of pigment, &c.), we can hardly doubt the generally accepted and more popular view, that the children tend to the mean of their parents; a view supported by Galton's admirable researches. If this be true sexual conjugation tends to diminish variations, and the author suggests that this is indeed the role of sex in evolution. Our attention has been so much engrossed with the changes of the environment, and with the consequent production and establishment of new varieties, that we have perhaps considered too little the fact that often for long periods of time the environment may remain stationary, and that under these conditions existing types must remain stable. The persistently

preserved types in our inland ponds and lakes, types such as the housefly and cockroach, which for ages have remained the same, illustrate this stability. Living matter must therefore be capable both of the power to vary and of stability; the first it possesses, the second it gains by sexual conjugation, which tends to prevent the slight deviations of a form which has become adapted to its environment from producing still further deviation by blending them together so that some, at any rate, of their progeny may preserve the useful ancestral qualities. To sex we owe our fairly defined species and genera; without sex we cannot doubt that life would exist in the form of innumerable varieties that we should fail to group together.

5. On the Relation of Mimetic Characters to the Original Form. By F. A. DIXEY, M.A., M.D., Fellow of Wadham College, Oxford.

An objection that has been often brought against the theory of mimicry, as enunciated by Bates and accepted by Darwin, is the difficulty of imagining the first stages in the production of a mimetic pattern. Fritz Müller endeavoured to meet this objection by alleging that mimicry chiefly originated between forms that already bore considerable resemblance to each other. The main instance (that of Leptalis melia) on which he relied in order to prove his point was not well chosen, for there is reason to think that he was in error both in considering that it represented the ancestral form of Leptalis and in supposing that it was not protected by mimicry. Nevertheless, his contention is sound in so far as it emphasises the fact that the process of mimetic assimilation depends rather on the development of old than on the starting of new features.

An illustration of this principle is afforded by a comparison of the nonmimetic butterflies Pieris locusta and P. thaloe with the mimetic species of the closely allied genus Mylothris, and with Heliconius numata, which serves as the model for the latter; all these forms inhabiting the same part of the neotropical region. An almost perfect transition can be traced on the undersides from the non-mimetic species of Pieris, through M. lypera 3, M. lorena ♂, M. pyrrha ♂, M. lorena, to M. pyrrha, this last butterfly being a very close copy of Heliconius numata. The whole series shows (1) that a practically perfect mimetic pattern can be evolved by gradual and easy stages without any violence or abruptness of change; (2) that it is not necessary that the forms between which mimicry originates should possess considerable initial resemblance; (3) that so small a beginning as the basal red patches on the underside of the hind-wing of many Pierines gives sufficient material for the assimilative process to work on.

The feebler development of the mimetic pattern in the males of this group calls for some explanation. No doubt the females require more protection, but does there exist any active check on the fuller assumption of mimetic patterns by the males? The retention of the original white by the latter sex has been in similar instances attributed to female choice; Mr. Wallace, on the other hand, thinks it due to the difference of habits in the two sexes, the females alone flying in company with the mimicked Heliconii. But this leaves unexplained the presence of a partial mimetic pattern in the male. The probability is that, although on the wing it may be advantageous rather than otherwise to the male, as Mr. Wallace thinks, to be taken for an ordinary white butterfly, yet when the insect is at rest and settled with the wings erect, any Heliconine resemblance would be to some extent protective; and the whole aspect of these males, the underside alone of which shows any mimetic features, is the resultant of these two divergent tendencies. The mimetic features of the male cannot be regarded as a mere incidental result of the more complete transformation of the female, because in many species of other groups the female is completely mimetic while the male shows no approach whatever to a mimetic change; moreover, there is a species of Hesperocharis (H. hirlanda) in which not only the male but both sexes show a partial mimetic pattern no further advanced than that of M. lorena & or M. pyrrha . It is difficult to believe that in this case the pattern is not in some degree protective.

Jenaisch. Zeitschr., vol. x.