raised by the shot be ignited so readily and carry on the flame so rapidly that it may produce explosive effects of a similar character to those caused by a gas explosion. The flame as it rushes along, if fed by freshly raised dust, may extend under these circumstances to very considerable distances, with results resembling, in their disastrous nature, those of explosions originating with, and mainly due to, fire-damp. This conclusion is very greatly strengthened by the evidence which the Messrs. Atkinson have brought together in the book before us. Their work indeed constitutes the most formidable indictment against coal-dust as a cause of colliery explosions which has yet been drawn up. In their capacity of Inspectors they have investigated with the most patient care the circumstances connected with what we may call six typical explosions. These were- All the explosions with the exception of that at Whitehaven were in the county of Durham. It would be quite impossible in the space at our disposal to follow the successive steps in the minute analysis to which the authors have subjected each of these explosions. We should require, moreover, many of the numerous plans of the colliery workings with which the book is illustrated were we to attempt such a task. All that can now be done is to point out the characteristic features of the several explosions, and to indicate the general conclusions which the authors draw from the consideration of the various circu nstances connected with them. We are conscious that in some respects this method of treating their work hardly does justice to the authors. It fails to convey any idea of the thoroughly scientific manner in which the Messrs. Atkinson's investigations have been conducted; of the minute and painstaking mode of their observation; or of the care and skill with which their deductions have been made. The authors, even in the earlier pages of the book, make their position in regard to the question of Gas versus Dust perfectly clear, but not even the most prejudiced opponent of the dust hypothesis can complain of the manner in which the evidence is presented. The Durham explosions presented many features in common. In the first place no accumulations of gas were known to exist in quantity sufficient to cause the widespread destruction which happened, nor were such accumulations considered possible. In all these explosions the downcast shafts were more or less damaged. At Trimdon Grange, Tudhoe, West Stanley, and Usworth the explosions did not cross the downcast shafts; these were wet, and the roadways near them were damp. At Seaham the shaft was dry, and the explosion crossed it and extended far beyond it. In all cases the violence and flame of the explosions were confined to roads on which there was much coal-dust. The explosions were most violent in the intake and haulage roads, or between the downcast shafts and lamp-stations, ie. in places where practically no gas was to be expected, and where naked lights were in constant use. The path of the explosion was in all cases that of the fresh air traversing the pit : in no case did it extend by means of the return air-way The return air-ways carry off the gases evolved in the pit, but are practically free from dust. In certain of the intake air-ways at Seaham and Usworth no coals were led, and they were consequently comparatively free from coal-dust: no traces of the explosions were observed in these roads. The explosions were in many cases arrested where the haulage roads were wet. In no instance did the explosion ascend or descend vertically through staples or shafts communicating with other planes of workings. If the explosions were due to gas, their extension would not be influenced by the direction of a communicating passage; on the other hand, very little coal-dust collects in vertical passages. In almost every case of an explosion which could with certainty be attributed to fire-damp, there is evidence that men have been alarmed and have attempted to escape from the workings before the actual occurrence of the disaster: in all the five Durham explosions there was no indication that any movements had taken place amongst the men suggestive of alarm; their bodies were found in the places where their work required them to be, close to their tools and lamps. At Seaham, Tudhoe, West Stanley, and Usworth the explosions were simultaneous with the firing of shots in stone; in these cases the explosions occurred when the pits were occupied by stonemen and repairers and at the only time when the operations of the mines allowed the firing of shots. At Seaham, Tudhoe, and Usworth the shots were fired on a main intake air-road and at points where currents of air of between 20,000 and 30,000 cubic feet per minute were passing. At West Stanley the shot was fired, in stone, at a working place by a naked flame, and the air in the vicinity would probably contain a small quantity of fire-damp, but not sufficient in amount to show its presence in the safety-lamp or by itself to be explosive. In the other cases it is almost impossible to conceive that the air could contain any sensible quantity of gas. At Seaham it would be necessary to assume that the gas came down the shaft, or that there were three separate and simultaneous outbursts of it on the three main roads diverging from the shaft. At Tudhoe, where the air came direct from the surface by two shafts, it would be necessary to assume two separate and simultaneous outbursts. At Usworth the air had passed no working place, and could hardly have contained even a trace of fire-damp. At West Stanley no appreciable quantity of gas could be present in the main intakes, although a small quantity might be contained in the air near the place where the shot was fired. There remains the Trimdon Grange explosion, which, was unconnected with shot-firing. There was distinct evidence that it originated with the ignition of gas at the light of a boy engaged at a pump in connection with some drowned workings from which gas was found to issue and that it extended with great force to parts of the pit more than a mile distant from its origin along the main intake air-ways. Now all the circumstances connected with the Durham disasters make it almost certain that the main agent in the propagation of the explosion was dust, and in three ་ ན out of the five cases it was dust alone. In four out of the five cases the immediate cause was shot-firing, but in no instance was the shot blown out. It is not at all necessary that the shot should be blown out to cause the ignition of the dust-cloud which the concussion raises in a dusty road. Properly fired shots show flame even when they dislodge the stone or coal; and the flame is often considerable if there has been an overcharge of powder, or if small coal or earth mixed with coal-dust has been used, as frequently happens, in the tamping. At Seaham, Tudhoe, West Stanley, and Usworth the flame of the shot ignited the dry inflammable dust dislodged from the roof or raised from the floor by the concussion of air which followed, and the explosion was propagated by fresh dust-clouds raised in the manner described by the Royal Commissioners. At Trimdon Grange an explosion of fire-damp operated in the same way: the violent movement of air resulting from the ignition of fire-damp and air raised a cloud of coal-dust into which the flame from the fire-damp passed, and the ignition of the coal-dust propagated itself as in the other cases, and, as in these, continued so long as it was fed by fresh fuel. This rapid ignition of dust containing upwards of 80 per cent. of carbon would result in the formation of large quantities of carbonic acid, and possibly even of the more poisonous carbonic oxide: when it is considered that it is impossible to live in air containing even 31 per cent. of carbonic acid, the deadly character of the after-damp so formed is readily conceivable. THE HE first edition of "Primitive Marriage" appeared in 1865, and the book was already extremely rare when, in 1876, it was reprinted as the first part of the "Studies in Ancient History." The reprint also soon became scarce, and while the influence of the author has been steadily growing, and almost all students of early society have come to attach great importance to his speculations, his principal writings have for some years been almost inaccessible. This new edition therefore supplies a real want, and it is doubly welcome for the sparing, but judicious, notes and appendixes which the editor, Mr. D. McLennan, has attached to his brother's book. "Primitive Marriage" broke ground in a new field of research, and, as the point of view was wholly novel, the collection, sifting, and marshalling of the evidence on which the argument was based was entirely pioneer's work. At the close of his life, McLennan was in possession of a much larger material; he had pursued his argument in new directions and to further conclusions, and on one or two points he had come to change his views. But new research had only confirmed the main lines of the argument sketched with so firm a hand in his original essay; and read with the caveats which his brother has introduced at one or two points-chiefly as regards the interpretation of the Levirate, and the prevalence of Agnation-the present reprint may be taken as generally representing, so far as it goes, the author's final conclusions on the subjects discussed. I say so far as it goes, for in many directions his conclusions had been added to and his views developed. The editor promises us a second volume, to consist for the most part of writings hitherto unpublished, which will throw a good deal of light on these new developments; meanwhile he has restricted himself in the notes "to certain matters on which the author had announced a change of view, and to certain others where circumstances had made an additional statement imperative." Of the additional statements, the most important is contained in two long notes appended to the essay on Morgan's "classificatory system" of relationships, in which it is clearly made out that Morgan's theory rests on misconception of the facts, and that the supposed classificatory system of relationship is not a system of relationship at all, but a system of terms of ceremonial or friendly address, used in conversation even between persons who are not related to one another in any way. This comes out so clearly in the cases about which we are best informed, that it is very questionable In striking contrast to the Durham explosions was that at Whitehaven. This was in a wet pit; the coal being worked was wet, and all the surroundings were damp, and free from dust. The cause of the explosion was gas, which was known to be in the pit, and frequently present in large quantities. Although it is probable that some 30,000 cubic feet of an inflammable mixture of air and fire-damp were ignited, the explosion was confined to a limited area of the workings, which extend to nearly three miles from the shafts. Seven men were within the district of the explosion, of whom three escaped. The survivors stated that all the men were alarmed by the appearance of gas immediately before the explosion, and hurried away. In the act of retreating the gas ignited at a lamp which was afterwards proved to have been defective and to allow of the passage of the flame. This the authors say was the most considerable explosion of fire-damp and air that they are acquainted with. They have personally investigated during the last twelve years almost all the explosions occurring in the North of England, and they cannot point to a case where there was direct evidence of so large a quantity of fire-damp and air exploding. The moral of all this is obvious. It can scarcely be gainsaid that some of the most disastrous explosions of the last thirty years are primarily to be attributed to the practice of firing gunpowder in dusty mines. That under certain circumstances gunpowder can be used with safety is allowed. But the Royal Commissioners have issued a warning in no uncertain terms. They have convinced themselves that the abolition of the use of powder in dry and dusty mines will not generally involve any formidable inconvenience, inasmuch as the work which is accomplished by its employment both in coal and in stone can now be performed with equal efficiency, and at very little whether the facts so laboriously collected by Mr. Morgan can be used to throw light on the early history of the family. From his plan of reprinting the book as it stood, with no more annotation than was absolutely necessary, the editor has departed only in one point. The appendix containing "additional examples of the form of capture" has been re-cast and enlarged upon the basis of a paper of J. F. McLennan published in the Argosy in 1866, but with additions from other and more accurate sources. The reasons for adopting this course are obvious: the new matter in this appendix could not conveniently have been reserved for the promised second volume, and the facts are so arranged and explained as to confirm the author's argument, and effectually dispose of the notion that the form of capture in marriage is to be explained by maidenly bashfulness. It will be seen from this brief account that, sparing as the editor's additions are, they make the new edition of the "Studies" well worthy of the attention of those who already possess the book in its older form. And to the not smal class of students of early society who know McLennan's work only at second hand or by one hasty perusal, it may not be unprofitable to say that this is emphatically a book of which a general knowledge is not sufficient, inasmuch as some of the most important and interesting points are precisely those which are almost sure to be missed on a first reading. For this, perhaps, McLennan himself is partly responsible, for in giving to "Primitive Marriage" the subordinate title "an inquiry into the origin of the form of capture in marriage ceremonies," he seems to fix attention on what is only the starting-point of a far-reaching research. In print and in conversation one often meets with the notion that the doctrines of marriage by capture and kinship through women only are mere archæological curiosa, and that for the study of later law and custom it is quite indifferent whether these things are true, or whether, on the contrary, mankind started from the first with male kinship. But the importance of McLennan's researches lies largely in the demonstration that the structure of society under a system of kinship in the male line which has been preceded by kinship through women cannot be the same as would be reached by a race which has had male kinship from the first. Other writers have taught a doctrine of the priority of kinship through women, but no one except McLennan has accurately developed the consequences of the doctrine, and shown how it solves a problem which, though ignored by most writers, is of the highest importance, namely, the origin of gentes within a nation. Like all really original thinkers, McLennan has for one of his chief merits that he recognised the existence of difficult problems in matters which ordinary people pass over without seeing any difficulty at all. And therefore precisely those passages in his writings which on a hasty reading seem needlessly laboured and proper to be skipped are found upon re-perusal to be particularly useful and stimulating. A word may be said in conclusion on what is promised for the second volume. It is satisfactory to know (p. 75) that it will include a short essay on the origin of exogamy. And from a note at p. 176 it may be inferred that in this essay the origin of exogamy will be sought in a state of custom. society where marriage by capture was an established We are also promised (p. 63) an essay on the marriage law of the Australian Kamiraloi, one of those highly complex problems in which McLennan's powers of analysis ought to appear at their best. From notes on pp. 109 and 228 it appears that part at least of McLennan's hitherto uncollected essays in the Fortnightly Review, including the papers on Totemism, or "On the Worship of Plants and Animals" (1869-70), will also be republished. It is to be hoped that in these reprints the editor will allow himself, in one direction, greater freedom of annotation than in the present volume. The Totem papers are in some respects the least finished of McLennan's writings, the evidence of totemism in the nations of ancient civilisation being much too largely drawn from second-hand sources. This gives an appearance of weakness to the whole structure of the argument, which has been very prejudicial to the influence of a most original and striking investigation. In point of fact a few of the detailed pieces of evidence ought to be abandoned altogether, but enough remains to leave the substance of the argument unaffected, and this ought to be clearly brought out by notes, referring to original authorities of unquestioned reputation, or giving up statements that cannot be authenticated. Even in the present volume one misses some notes of this kind. The polyandria of the Athenians (p. 235) rests on better evidence than the | story which Augustine cites from Varro (Clearchus ap. Athen. xiii. p. 556 d.). Again, the note at p. 47, in which an attempt is made to prove the existence of the form of capture among the Hebrews from the phrase "to take a wife," ought rather to have been withdrawn than again built upon by the editor at p. 181; and what is said of the marriages of the Persians at p. 219 sq. requires careful revision. W. ROBERTSON SMITH BRITISH HYMENOMYCETES WE are glad to welcome this second volume so It is of considerable importance to students that a work which professes to include all British species, up to date, should satisfy all reasonable expectations. The first volume omitted some forty species, and the present is by no means perfect. We open at p. 232, and find under the genus Solenia one solitary British species recorded, that of Solenia ochracea. Surely our author could not have been ignorant of the fact that Solenia anomala, P., is still more common, and was recorded by Berkeley in the "English Flora" (p. 199) fully fifty years ago. Neither could he have forgotten that another species was included in Cooke's "Hand-book" (p. 329) under the name of S. candida, since corrected to S. fasciculata. As these specimens were collected near Batheaston, by no other than Mr. C. E. Broome, and confirmed by the Rev. M. J. Berkeley, no doubt can be entertained of their being authentic. Furthermore, the name was corrected and the species figured by Berkeley and Broome in the Annals of Natural History, December 1870, No. 1301. The fourth species is Solenia stipitata, Fuckel, of which there are specimens in the Kew Herbarium. It cannot be conceded that a "Flora" satisfies all reasonable expectations when in one genus only one of four species is recorded. Turning to an allied genus, that of Cyphella, we seek in vain for C. Curreyi or C. albo-violascens (which may be identical), C. cyclas, Cke. and Phil., C. punctiformis, Fries, C. villosa, Pers., all but one of which are well-known and widely-distributed species. Whether the species under the genera Stereum and Corticium might have been arranged in a manner more in accordance with modern ideas, and far more useful to the student in their identification, may be left an open question. Those who are not facile in the use of the microscope may find it convenient to follow Fries, who paid little attention to microscopical characters, but surely in a large and difficult genus, such as Corticium, no assistance should be despised. We observe, with some surprise, the genus Microcera, of Desmazières, included in a work devoted to British Hymenomycetes (p. 308) with the intimation "no British species." The fact is that Microcera coccophila, Desm., which is the type of the genus, has been found in Britain, and is recorded on p. 556 of Cooke's "Hand-book," and furthermore it is also true that it is not a Hymenomycete at all, but the conidia of one of the Sphæriacei, and is included as such in Saccardo's "Sylloge Fungorum" (vol. ii. p. 513). This singular double error might have been avoided had some mycologist been consulted who had not confined his attention exclusively to the Hymenomycetes. them as varieties of one species. At any rate, there is no good reason why such closely-allied forms should be separated by four-and-twenty intermediate species. The mention of localities for species throughout the work is so vague, that some explanation should have been offered. When only one locality is given, the inference which would be drawn by the majority of readers would be that no other British locality was known at the time for that particular species. That this conclusion would be wrong is manifest from Hydnum Weinmanni (p. 242), which may be taken as an example. The locality cited is "Bristol," but Bristol is not the only, or the most important station for this species in Britain, because it occurs plentifully in the neighbourhood of Carlisle. If the intention was simply to indicate the locality where the species was first found in these islands, then again we fancy it is inaccurate, because, as we believe, it was first discovered by the late Rev. A. Bloxam, at Gopsall. The only solution we could suggest is that "Bristol" is the locality mentioned in Berk-ley's "Outlines," and it was accepted as the only authentic record, without inquiry. Some species are stated to be "common," others "frequent," and others "rare," and when, in the absence of any one of these terms, a single locality is given, it is a fair inference that only one locality was known to our author, and that was the reason why it was given. Assuming this to be the case, we fancy that a very large number of these single localities could be challenged as not unique. In addition to a Glossary" of five pages, we are glad to find a good index of genera and species, but we search in vain for any clue to the contractions, in some cases only a single letter, employed in quoting authorities. Under nearly every species follows a line or two, sometimes five or six lines, of hieroglyphics, to which figures are appended. It may be all clear enough to the Rev. John Stevenson what is intended to be conveyed by "Quel. t. 11, f. 1," or "Viv. t. 27," or "C. Illust., pl. 276," but who these illustrious persons are, or what they have done, to be curtailed in such wise, is nowhere indicated. Surely the author must have determined upon giving a key to these mysteries when he first commenced to employ them, and, in the hurry to issue the second volume, quite forgot the "students," even if he remembered the "scholars," and closed the book before he had finished his work. A summary of the contents of these volumes, as they stand, exhibits the following results as compared with the last preceding work on the same subject :"Hand-book of British Fungi" Stevenson's "British Fungi ... ... 1044 1675 The limits of species is another open question, and it is armeniacus, Berk. (p. 215), is generally admitted to be only a resupinate condition of P. amorphus, Fries, and should not be continued as a distinct species. P. Herbergii, Rost (p. 195), is placed as an ally of P. sulphureus in the section "Caseosi," whereas P. cuticularis is found (at p. 202) in "Spongiosi." Unfortunately for this arrangement, the two species (P. Herbergii and P. cuticularis) are so closely allied that sometimes it is difficult to distinguish the one from the other, except by the difference in size of the pores, and hence some regard ... 699 1183 or, an addition of 484 species, leaving only 147 species There can be no doubt that all that portion of the work which contains translations from the "Monographia" of Fries will be exceedingly valuable to British mycologists, and this extends through the whole of the first volume and 165 pages into the second; the only regret being that the few remaining species, which have not as yet been recorded in these Isles, were not inserted in brackets, or published as an appendix, so that the whole of Fries's excellent work might have been in the hands of every mycologist in this country. Perhaps even now such an appendix might be published, and no doubt it would meet with a hearty welcome. Despite of such strictures as we have been impelled to make, we venture to hope that the present edition will soon be exhausted, and that its author will be called upon to prepare a new and revised edition, with a key to all the mysteries of the old one. THE OCEAN M. C. C. (Leipzig und Prag: treated of. After having made known the principal phenomena regarding the temperature of the ocean and its distribution, Krümmel treats of the glacial phenomena of the Arctic and Antarctic Oceans, pointing out the limits of the floating ice and icebergs in each region, and the influence of these regions on the questions of general oceanic circulation. The last chapter is reserved for a consideration of the movements of water, such as currents, waves, and tides. Such is the general order and method of this manual. There is no attempt to give any general notion of the life of the shores, deep sea, and surface of the ocean, or of any of the phenomena due to organisms. The author shows himself to be everywhere au courant with the most recent discoveries in his subject. It would appear, however, that he has not had an opportunity of consulting the "Narrative of the Cruise of the Challenger," published last year, or he would have embraced in his descriptions some additional interesting details and general views. The work is illustrated by many woodcuts and small charts, some of which are instructive, others conveying little information to the reader, but THE HE great interest which oceanographical studies when the low price of the book (one shilling) is rememhave aroused within the last few years is shown in bered it would be unfair to criticise closely these illusa marked manner by the publications destined to popu-trations. Dr. Krümmel has attained the object he had larise the notions acquired respecting this vast and important chapter of physical geography. Not long after the appearance of the "Lehrbuch der Ozeanographie" by Boguslawski, whose untimely death has interrupted the publication of the second volume, we have a new and small manual by Dr. Otto Krümmel, whose name is already known to oceanographers. Der Ozean. Von Otto Krümmel. This little treatise is clearly written, and the most important general notions concerning the physical geography of the sea are well stated, and discussed with ability. The author has succeeded in expressing briefly the essential notions about the ocean, which have been recently acquired by the Challenger and other deep-sea expeditions. The author describes, in the first place, the ocean's surface and its subdivisions ("Die Meeresflächen und ihre Gliederung "); discusses the relation of oceanic and terrestrial areas from the point of view of their respective size; indicates the distribution according to hemispheres; and points out the classification he has adopted into oceans, properly so called, with their general systems of ocean currents, and secondary seas, which are more or less cut off from the great oceans. The secondary seas are again subdivided into interior, or inter-continental, and border seas, situated on the outer edges of the continents. The volume of oceanic water is then estimated. In the second chapter the interesting questions connected with the deformation of the level and surface of the ocean, owing to the attraction of the continental masses, are examined The depths and contours of the ocean basins are next pointed out, and the work of the Challenger and other deep-sea expeditions, together with the apparatus employed, is described. The observations of the Challenger upon the nature and distribution of deep-sea deposits are summarised. The physical and chemical properties of sea-water are set forth in a special chapterthe salinity of the ocean, its distribution and origin; the gas contents; the transparency and colour, are, in turn, in view-to popularise in a scientific manner our know- LETTERS TO THE EDITOR [The Editor does not hold himself responsible for opinions expressed by his correspondents. Neither can he undertake to return, or to correspond with the writers of, rejected manu[The Editor urgently requests correspondents to keep their letters scripts. No notice is taken of anonymous communications. as short as possible. The pressure on his space is so great that it is impossible otherwise to insure the appearance even of communications containing interesting and novel facts.] On the Connection between Chemical Constitution and Physiological Action As regards Dr. Brunton's letter in last week's NATURE (p. 617), I would express myself as more than satisfied with the personal explanation, but Dr. Brunton has not noticed the mot important point to which I wished to call attention, viz. that whatever may be the value of my experiments, as showing a connection between physiological action and chemical constitution, the researches of Crum Brown and Fraser have really no bearing on the subject, for the simple reason that they had no knowledge of the chemical constitution of the re-agents they employed. There is an old receipt for cooking a hare which commences "First catch your hare," and in attempting to show the influence of change in chemical constitution on phy iological action, it is well first to get a constitution. In the last edition of Watts's "Organic Chemistry" (1886) it is stated, "All these bases (the alkaloids), like the amines, are derivatives of ammonia, but their molecular structure is for the most part unknown." Even as regards inorganic compounds, our knowledge of their chemical constitution is not the most definite, but I believe that the arrangement of the elements in isomorphous groups expresses most clearly the resemblance in the chemical constitution of their compounds. confess that I cannot find anything showing the connection be After again reading carefully Dr. Brunton's paper, I must tween chemical constitution and physiological action, except, perhaps, in the case of the alcohols. Here we have a class of bodies in which the different members of the series have probably the same relation to each other as the elements in the same isomorphous group, and it is an interesting fact that not |