believe in the old men rather than in the new." It is true that in these cases he is referring to special points, but again and again one cannot but feel in reading the book that the writer pays undue regard for authority, without considering that his "old masters were not acquainted with all the facts which we now possess, and that they themselves changed their views. Sedgwick, for instance, came to believe in an Ice age. Moreover, if these were old masters, so were Hutton and Playfair, Lyell and Buckland, whose views are not always so palatable to the author. It may be remarked, also, that a knowledge of mathematics and physics was not confined to the geologists of those days. One of the most ardent of the existing advocates of ice-erosion, concerning whose paper on ice action in Skye (Trans. Roy. Soc. Edin., vol. xl., 1901) Sir Henry is silent in these two volumes, was a high wrangler, and took a first class in physics at Cambridge.

The theory of an Ice age was largely put forward owing to the existence of rounded and striated rocksurfaces and scratched and polished boulders. These resemble similar productions of modern ice to such a degree that the geologist has no more hesitation in referring them to ice-action than he has to assign the formation of the pebbles of a river to streamaction. The inference drawn from the existence of these phenomena has been supported by a host of other observations, biological as well as physical, and if Sir Henry should succeed in disproving the existence of an Ice age he will also break down the essential principle of geology, "that like effects imply like causes.

It would be impossible in a brief article to discuss all the questions raised in this part of the work. We must content ourselves with a few observations. Though reference is made now and again to the Greenland ice and to the ice masses of Spitsbergen, it is the glaciers of the alpine type to which most frequent appeal is made. To this we shall recur, but in the meanwhile would invite the author's attention to yet another treatise concerning which he is silent, where another type of ice work is described, namely, I. C. Russell's volume on the Malaspina Glacier (thirteenth annual report of the U.S. Geological Survey).

When describing the Till or Boulder-clay, the author quotes a description of it by Prof. James Geikie, and goes on to observe, "this being without question the most typical of so-called glacial deposit, it is a remarkable fact that no such deposit is now being made, so far as we know, by land-ice anywhere." He must have overlooked a passage in a paper to which he elsewhere refers, by Messrs. Garwood and Gregory, on the glacial geology of Spitsbergen (Quart. Journ. Geol. Soc., vol. liv.). They say:

"On the broad plain at the foot of Booming Glacier we found some square miles of a tough mud containing boulders and pebbles; it only needed to be dried and hardened to form an ideal Boulder Clay. Clearly this deposit had been laid down by landice."

The author objects to the sharp line which is drawn by many geologists to show the margin of the ice at its period of maximum extension, and denies the existence of any evidence for this, arguing that the Boulder-clay, the masses of gravel and loam, and the loess are genetically connected. Of this we shall doubtless hear more when the third volume appears.

Much is naturally made of the conflict of opinion among geologists concerning the occurrence of interGlacial periods, and the relative importance of landice and floating-ice in producing the phenomena generally taken to indicate the occurrence of a Glacial period. These questions are certainly not settled to everyone's satisfaction, but they in no way invalidate the conclusions which have been drawn as to the existence of an Ice age.

Though we do not agree with the author in his main conclusions put forward in this section of the work, we must admit that much that he writes is worthy of consideration, even though his views seem exaggerated. For instance, he argues that much of the material forming the drifts was broken up prior to the so-called Ice age, and this we believe to be true, even though the breakage did not occur in the manner advocated elsewhere by the author; but if true, it invalidates the appeal to modern Alpine glaciers to prove the inadequacy of ice as an erosive agent. The loose materials ready to hand at the beginning of Glacial times would supply the ice with the tools for rasping and grinding. As that material became comminuted, unless new material was supplied in abundance, the ice would become less effective as an eroder. Also ice, like water, has a base line of erosion beneath which it cannot work. This line may have been reached in the case of Alpine glaciers, and the supply of material to the sole have been also largely diminished, in which case one can no more argue from what Alpine glaciers are now doing as to the effects of land ice in the Glacial period than one can explain the cañons of the Colorado by reference to a little stream which has established its base level.

Throughout the work much has been made of the conflicting views of geologists as to the details of ice action. Sir Henry is obviously greatly impressed with the fact that in the long and arduous attempt to unravel the Gordian knot the skein sometimes seems to have become hopelessly twisted; but he who carefully studies the process of disentanglement sees that, notwithstanding the many kinks, the tangle is becoming less. The author, impatient of the slow process, has elsewhere attempted to cut the knot, and will evidently give reasons for this act in the third volume. We fear that the attempt will not be regarded as successful, either by the "ultra-glacialists or by geologists in general.

We cannot recommend the book to geological babes and sucklings, but it will well repay perusal by the advanced reader. He will forgive the "energetic adjectives and adverbs," which are hardly necessary to a calm and dispassionate discussion, on reading the author's frank apology in the preface. The store of facts collected in the book is of the utmost value

to the student of glacial geology, though we wish that references to the original memoirs had been in all cases added. There are, as we have tried to show, many valuable criticisms and suggestions contained in the work. Lastly, it will prove a useful intellectual exercise to weigh the author's arguments in the balance. For these reasons we believe that readers who have an extensive acquaintance with the facts and principles of geology will read the book with profit-and with pleasure. J. E. M.

PHYSICAL CHEMISTRY. Theoretical Chemistry. By Prof. Walther Nernst. Revised in accordance with the fourth German edition. Pp. xxiv+771. (London: Macmillan and Co., Ltd., 1905.) 158. net.

THE

on

HE fact that three further editions of the German text of Nernst's well known treatise theoretical chemistry have been called for since the appearance of the original in 1893, affords ample testimony to its intrinsic merits. An English translation of the first edition by Prof. C. S. Palmer appeared in 1895, and this, until now, has been the only English version.

During the last ten years much valuable work has been carried out in the province of physical chemistry, and the publishers have recognised the necessity of bringing the English edition up to date. With that object Dr. R. A. Lehfeldt has translated the whole of the new matter contained in the fourth German edition and has revised certain parts of the original translation.

It has been the reviewer's experience to hear the original translation adversely commented upon, and it is perhaps to be regretted that the bulk of the old text remains as it was in the first edition. After careful perusal of the work, it is indeed difficult to suppress the feeling that a better result would have been attained by an entirely new translation of the fourth German edition.

Two new chapters in the work under review deal with "The Atomistic Theory of Electricity and "The Metallic State." In the first of these an account is given of the electron theory and of the phenomena of ionisation and electric conduction in gases. In the second the nature of the metallic condition is discussed on the basis of results which have been obtained by the study of the freezing point curves and of the electrical conductivity of mixtures of metals. These chapters form very interesting reading, although, of course, it has not been possible within the compass of seventeen pages to give more than the briefest outline.

The space given to electro-chemistry has been extended from 26 to 46 pages, and the exposition of the subject-matter greatly improved. The application of thermodynamics and of the osmotic theory to electrochemical systems is now treated in separate chapters, and many new observations bearing on the theory of electrolysis have been incorporated.

It is not possible to mention more than a few of the alterations and additions which have been made

in the text generally. One notes with pleasure that the somewhat abstruse exposition of energy relationships in the introductory chapter has been made more lucid. The discovery of the inert gases of the argon series has led to much discussion of late years in reference to the periodic classification of the elements, and these recent views are summarised in the chapter on the atomic theory.

Other important new sections deal with Werner's theory of molecular compounds, catalysis, the mechanism of autoxidation processes, tautomerism, and the kinetics of heterogeneous systems. The view that tautomerism is due to the co-existence in dynamic equilibrium of mutually transformable isomeric substances seems to be very probable in the light of recent work. In this connection the interesting observations of Hantzsch on the transformation of the tautomeric forms of nitrophenylmethane and similar bodies are recorded, but one looks in vain for any reference to Lowry's investigations on dynamic isomerism. In reference to the kinetics of heterogeneous systems and the mechanism of chemical change, it is now recognised that many gaseous reactions, usually regarded as taking place in a single phase, are possibly examples of changes essentially conditioned by phenomena at a boundary surface. The rate at which arsine or phosphine decomposes is in accord with the formula for a unimolecular change, but this agreement really affords no conclusive argument with reference to the mechanism of the change. The measured rate of change has possibly nothing whatever to do with the chemical change involved, but merely with a physical change at the surface of the containing vessel. In a third edition reference should be made to this in the section dealing with the mechanism of reactions on pp. 562-564.

Of necessity, much new work has had to be left unmentioned in the new edition, but the author is to be congratulated on the large amount of new matter which he has been able to introduce without appreciable alteration in the size of the volume. With the issue of this second edition one may confidently anticipate that Nernst's book will still maintain its position as one of the classics of theoretical chemistry. H. M. D.

and in giving assistance to the investigations of others. The volume thus consists largely of additions and notes originally appended to memoirs by

other authors."

Hence, although we meet abundant evidence of Stokes's constant occupation with scientific subjects, and of the characteristic generosity with which he placed his powers at the service of others, we miss something of the more spontaneous activity which characterised his earlier period. We find various proofs, however, that the subjects which had first fascinated him were never long absent from his thoughts; and occasionally they receive a flash of unexpected illumination. We may cite the various notes on water-waves, the brilliant little paper on semi-convergent series, and the admirable interpretation of Prof. Hele-Shaw's experiments on the flow of a viscous liquid between parallel plates. We have also

a record of the keen interest which in the last few years of his life he took in the subject of Röntgen rays. The lecture (p. 256) which he gave to the Manchester Literary and Philosophical Society in 1896 was written out (with the help of reporters' notes) after delivery; bright and genial as it is, it gives no adequate idea of the buoyant freshness and vivacity which characterised the oral exposition.

The volume includes, by a happy determination, a collection of the papers set by Stokes in the mathematical tripos, and in the old Smith's Prize examination. It is well known that through this unusual channel several important scientific results were first made known to the world; for example, the notion of group-velocity, and the famous "Stokes's Theorem,' respecting which we have an interesting historical note by Prof. Larmor. We suspect that a mathematical antiquarian might make further interesting "finds." If we are not mistaken, we detect prior publications of a remarkable theorem relating to the infinite product for sin x, and of a definite integral property of Bessel's functions, which are usually

attributed to Weierstrass and to H. Weber respectively. Of course, no one, least of all Stokes himself, would attach much importance to the question of priority under these conditions; but such instances are of interest as showing, in unexpected directions, the singular vigour and independence of Stokes's mind.

The Royal Society obituary notice, with its authoritative appreciation of Stokes's scientific researches by one of his keenest admirers and disciples, forms a fitting accompaniment to this monumental publication. The volume is further adorned by an excellent photograph by Mrs. Myers, of date 1892.

The scientific world will await with great interest the publication of the "volume of biographical character, to be occupied in part by a selection from Sir George Stokes's voluminous scientific correspondence, including some unpublished manuscript material,' which is promised in the preface. The great energy with which Prof. Larmor has discharged his present honourable task justifies the hope that we shall not have to wait too long for the proposed supplement.

H. L.

OUR BOOK SHELF.

Notes on the Drawings for Sowerby's English Botany." By F. N. A. Garry. Reprinted from the Journal of Botany, 1904-5. Pp. 276. (London: West, Newman and Co., 1905.) Price 6s. THE series of volumes known as "English Botany" was begun in 1790 by James Sowerby, the botanic artist, who engaged Dr. James Edward Smith, the possessor of the Linnean collections and founder of the Linnean Society, to describe the plants depicted by him. At first the name of the draughtsman only appeared on the title-page, but in 1795 a preface to the fourth volume by Smith acknowledged his authorship, and he was much annoyed in after years by "the flippancy with which everybody quotes "Sowerby," whom they know merely as the delineator of these plates, without adverting to the information of the work, or the name of the author." The artist and ings of the phanerogams and vascular cryptogams, those who followed him preserved the original drawwhich ultimately came into the possession of the trustees of the British Museum, and are now in the department of botany. Here are to be found the drawings, with impressions from the original plates, by side on the same sheets. The drawings (which and also from the third recast edition, laid down side

had been submitted to Smith for his criticism and his text which accompanied them) bear many notes and directions to the engraver, which are of great interest as showing not only the state of botany at the time. but mentioning the numerous contributors of plants excellent service in the laborious task of transcribing to the work and its supplement. Mr. Garry has done and editing these notes, which can now be read by those who have not seen the originals themselves. Turning over these pages, the writer is reminded of the days when, more than thirty years ago, he first made acquaintance with the drawings in the old rooms of the department at Bloomsbury, recalling the charm they possess for all who care for the history of the native plants of Great Britain.

Without going into detail it would be impossible to set out many most interesting items which are to be found in the pages of this modest reprint from the made their first appearance. We have here glimpses two years' supplements to the journal in which they of a big book in the making, which extended in the first instance to thirty-six volumes and closed in 1814 with a general index. Further discoveries and greater discrimination of critical forms induced the beginning in its fifth volume; the text in these later volum of a supplement in 1831, which died out in 1866 was by many hands, amongst them the most active and critical of the botanists of the day. These last plates are now in the Fielding herbarium at Oxford, whence they were borrowed by the author so as to complete his work. B. D. J.

A Text-book of Chemical Arithmetic. By Horace L. Wells. Pp. vii+ 166. (New York: John Wiley and Sons; London: Chapman and Hall, Ltd., 1905.) Price 5s. 6d. net.

In the preface it is stated that this book "is designed especially for the use of students of quantitative analysis, many of whom, even after having taken extensive courses in higher mathematics, shou little ability to solve simple chemical problems. Certain portions of the work are suitable also for the use of those who are studying elementary chemistry." It appears, therefore, that an American professor is no better off than his English cousin in this matter of student arithmetic The difficulty is two-fold. In

the first place, the student has never been taught arithmetic in relation to actual measurements, but has been exercised in fictitious transactions with oranges and nuts, rods, poles or perches, and vats into which liquor flows at the rate of so many gallons a minute and out of which it flows (notwithstanding the dwindling pressure) at another exact and steady rate. The result is that the student has no idea of the relation of magnitude to measurement, and no opinion whatever on the subject of significant figures; he cannot use logarithms or a slide-rule, and is unpractised in contracted methods of computation. In the second place, it is very likely that he has no sound idea of proportion. Given a student in this condition -and it is still the common case the teaching of what is called chemical arithmetic becomes a serious part of the duties of a teacher of chemistry. The fundamental numbers of chemistry-the atomic weights are proportional numbers, and it may be said without exaggeration that the failure to realise this and the inability to see how proportional numbers may be used for the calculation of absolute weights, locate the real pons asinorum of elementary chemistry.

In these circumstances any well considered attempt to expound the elements of chemical arithmetic is to be welcomed, and Prof. Wells has certainly succeeded in writing something on the subject which is likely to be very useful. He does not quite descend to the meanest capacity, but he deals in a very clear way with the meaning of figures and the limits of accuracy in measurement and computation. He also gives a good survey of the chief types of chemical problems, including all kinds of analyses and the corrections of gas volumes. Great pains are taken to impress the student with the importance of using common sense and judgment whilst performing arithmetical operations, and to this end set rules and stereotyped formulæ are avoided. An appendix to the book contains tables, including a well printed set of five-figure logarithms. Altogether the work is one that may be warmly recommended to the notice of English teachers. A. S.

The Physics and Chemistry of Mining. By T. H. Byrom. Pp. xii+ 160. (London: Crosby Lockwood and Son, 1905.) Price 38. 6d. net. THIS elementary class-book supplies information required for such examinations as the Board of Education principles of mining, stage i. The idea is a good one, as the principles of pure science upon which mining practice is based are apt to receive scant attention in mining classes. The author, who is chemist to an important colliery company, has, as lecturer at the Wigan Technical College, become acquainted with the needs of students, and he gives in concise form much useful information regarding the atmosphere, the laws relating to the behaviour of gases, the diffusion of gases, the composition of the atmosphere, water, carbon, fire-damp, combustion, coal dust, explosives, the composition of coals, the analysis of coal, the strata adjoining the Coalmeasures, magnetism and electricity. The language is simple, and chemical symbols are sparingly used. There is, however, a want of uniformity in nomenclature that might confuse the beginner. The terms "carbonate of magnesium" (p. 96) and "magnesia carbonate" (p. 125), “iron oxide and aluminia 46) and "iron peroxide and alumina" (p. 125) are examples. The author, too, should not have included Cumberland hæmatite among the ironstones, nor granite among the strata adjoining the Coal

measures.

(p.

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 manuscripts intended for this or any other part of NATURE. No notice is taken of anonymous communications.]

On the Absorption Spectrum of Benzene in the UltraViolet Region.

IN the Transactions of the Chemical Society for August Messrs. Baly and Collie, referring to the previous work of Baly and Desch (Trans. Chem. Soc., 1904, lxxxv., 1029, and 1905, lxxxvii., 706) on the absorption spectrum of acetylacetone and its derivatives and the conclusions arrived at, namely, that the absorption band is caused by dynamic isomerism, or rather isodynamic changes, are led to infer from the occurrence of bands in the spectrum of benzene that these also are caused by the making and breaking of the carbon bonds in the molecule of the substance. I have given a similar, but not identical, explanation of the cause of the bands in the spectra of uric acid, murexide, and the ureides, and have pointed out that there is but little difficulty in accepting a like explanation in order to account for the bands in aromatic hydrocarbons, seeing that this would harmonise with Kekulé's view of the constitution of benzene. The particulars are contained in two papers communicated to the Chemical Society on May 17, but as they are still unpublished I cannot refer to them in detail.

Messrs. Baly and Collie consider all the possible phases in change of linking between the six carbon atoms in benzene, and assign a band to each phase. In doing this they feel justified in assuming that an even number of carbon atoms is concerned in each individual process, and in accordance with chemical evidence it could scarcely be imagined otherwise. They argue that there are only seven different makings and breakings of bonds possible, to which seven different absorption bands should belong, and on investigating the spectrum of benzene they find only seven bands. Seven bands were photographed (Phil. Trans., 1879), as they remark, by Hartley and Huntington, but are given. The wave-lengths of lines in the ultra-violet had not been determined at that time (1878), with the exception of the principal lines of cadmium measured by Mascart, hence the reason for the absence of measurements.

по measurements

In a subsequent observation (Hartley and Dobbie, "Notes on the Absorption Spectrum of Benzene," Trans. Chem. Soc., 1898, lxxiii., 695) seven bands were photographed and measured, but one of these appeared to differ from the others in constitution, and it was indicated as doubtful; it is also a feeble band. The general character possessed by the first six bands was most distinctly marked in the four strongest; each was stronger and generally sharper towards the side where the rays of shorter wavelengths lie, and was weakened in the opposite direction, as if the bands were composed of groups of lines occurring closer together and being stronger towards the more refrangible edge. Baly and Collie appear to have overlooked some points of importance in this communication, since they

were

state that Hartley and Dobbie found only six bands, and that the measurements of the actual heads of the bands are not given. They give a series of numbers derived from Hartley and Dobbie's measurements which for comparison with their own are printed in a parallel column. The gist of the paper by Hartley and Dobbie was to show the structure of the benzene absorption spectrum partly by measurements and partly by the aid of a photograph. The bands which distinctly showed the structure numbered, but unfortunately the manner in which the Photograph was reproduced failed to render delicate details which were visible on the original plate. The statements contained in the paper appear, however, to have been clearly and fully understood by W. Friederichs, who photographed the vapour of benzene with a Rowland grating. He found fifty-six bands of absorption in its spectrum in the ultra-violet, which are arranged in eight groups, and he compared the principal lines of each group with the