seems to be generally believed by geologists that in natural conditions the melting-point of granite is more affected by water than that of basic rocks. A reconsideration of the data on this question is overdue; such a conflict of opinion on some of the most important experimental data of geophysics should not be allowed to persist." There is nothing in this, nor in the originals referred to, so far as I can find, to sustain Dr. Jeffreys's contention that the melting-point of rocks increases with the basicity; but quite the contrary. In confusing the melting-point with crystallizing-point serious error is involved. In the case of "crystallization-point" we deal with solutions: "The relation of the several compounds in a magma is one of mutual solution" (Harker, The Natural History of Igneous Rocks,' p. 169). It results" that the order in which the minerals crystallize is not determined by their relative fusibility as separately tested. This lowering of freezingpoint is the most characteristic property of solutions" (p. 170). The italics are Dr. Harker's. Again, volatile substances play an important part in determining the temperature of crystallization of the constituents; and as these substances escape more freely from extruded than from intruded magmas, the former may crystallize at higher temperatures than the latter. Also, volatile substances (such as water) are more abundant in acid than in basic magmas, and hence play a more important part in the crystallization of the former. "It follows that under like conditions acid magmas crystallize at lower temperatures than basic magmas. But there is no doubt as to the fact that the melting-point rises with the percentage of silica. Thus Barus and Iddings find as follows: :- These figures bring out strikingly the difference, already referred to, between the melting-point and the crystallizationpoint. Rhyolite is generally regarded as the volcanic equivalent of a granite. Clarke, who cites these results (Data of Geochemistry,' 1924, p. 301), continues: "At 1300° the basalt was quite fluid, but at 1700° the rhyolite was still viscid." Doelter found that basic rocks softened between 1000° and 1150°, and acidic rocks softened at 1200°-1300° (Phys. Chem. Min. 1905, p. 124). Clarke (loc. cit. p. 297) states the issue clearly. "In the geological interpretation of the melting-points there is one particularly dangerous source of error. We must not assume that the temperature at which a given oxide or silicate melts is the temperature at which a mineral can rystallize from a "magma." The experiment of floating a piece of granite in molten basalt may easily be made. The conditions are stable, because the crystallization temperature of basalt (1050°) is beneath the melting-point, or rather the decomposition-point, of orthoclase (1170°) and the melting-point of albite (1100°), and, of course, much below the rather indefinite meltingpoint of quartz. (See J. H. L. Vogt, 'Physical Chemistry of the Magmatic Differentiation of Igneous Rocks,' ii. 1926.) There remains one more point. Dr. Jeffreys refers to his deduction, on seismic evidence, of a low continental thickness, assuming that this has been verified. A recent paper by Mohorovičič in Gerlands 'Beiträge zur Geophysik' (Bd. xvii. Hept 2) disputes his low estimate, and contends that the thickness of the continental sial is about. 40 km. (loc. cit. p. 198). The discussion is still proceeding. Meanwhile the following statement by Dr. Jeffreys (Nature,' Sept. 25, 1926) referring to certain seismic determinations is of interest. He refers to the Kulpa Valley earthquake of 1909, the Wurtemberg earthquake of 1911, the Tauern earthquake of 1923, and the Oppau explosion. "The observations permit of a rough determination of the depths of the granite and basaltic layers. The former may be about 12 km., the latter about 20 km. Both are subject to an accidental error of about 4 km. addition there is a possibility of systematic error. Uncertainty as to the depth of focus may allow the thickness of the granite layer to be doubled. ... I think, therefore, that determinations of the depths of the layers by means of near earthquakes are not more reliable than those based on the Earth's thermal state, isostatic balance between continents and oceans, and the group velocities of surface-waves," etc. The italics are mine. In Isostatic considerations appear to favour 37 km. or thereabouts as the continental thickness (Nature,' Oct. 29, 1927). Thermal estimates based on surface gradients and the radioactivity of the rocks are not very trustworthy. The point is not very important, and does not directly affect the theory of thermal cycles. XXII. Double Excitation of Upper Levels in the Mercury Atoms by Collisions of the Second Kind. To the Editors of the Philosophical Magazine. GENTLEMEN, MAY AY we make some remarks concerning the very interesting paper of Professor R. W. Wood in the September issue of your Magazine on "Optical Excitation of Mercury, with Controlled Radiating States and Forbidden Lines (Phil. Mag. [7] iv. p. 466, 1927). Professor Wood describes the radiation of the mercury resonance lamp which contains some nitrogen. He finds "effects of obscure origin, which can be explained only with difficulty." Very strange is the development of the line 2856 caused by the addition of the nitrogen, whereas the resonance is only excited by light of shorter wave-length than 2750. The line 4916 from the same upper level 3 is also enormously strengthened. In observations made on the mercury resonance-tube with addition of sodium (Naturwiss. xv. p. 540, 1927) we have shown that upper levels of an atom are excited in high selectivity by collisions of the second kind if the energy of the colliding atom can be accepted very completely by the atom struck. At the collision of two metastable (in Wood's terminology 2p3) Hg atoms this effect of resonance can have the result that one accepts the whole energy, 2 x 4.68 volts, while the other returns to the normal state. This process has a certain probability, because nearly under the theoretical level of resonance (2 × 4·68=9·36 volts) an actual level of the mercury is located with 9.25 volts, that is, only 011 volt is to be transformed into energy of translation; this equals the thermal energy of the two atoms at 300° K. The level from which the mentioned lines 2856 (2p2-3S) and 4916 (1P-3S) originate is the level under the point of resonance 38. Investigations are under way aiming to substantiate this hypothesis of the accumulation of energy. Yours faithfully, H. BEUTLER and B. JOSEPHY. Kaiser Wilhelm-Institut für physikalisch XXIII. Notices respecting New Books. The Nature of the World of Man. By Sixteen Members of the Faculty of the University of Chicago. Edited by H. H. NEWMAN 8vo. Pp. xxiv+566, 132 figs. and 4 charts. (University of Chicago Press; Cambridge University Press, London, 1926. Price 20s. net.) THE purpose of this book is to present an outline of our know ledge of the physical and the biological world, and to show the position of man in the universe in which he lives. It contains the subject-matter of a "survey course given each year by its authors at the University of Chicago to a group of selected firstyear students of superior intelligence. The course was designed to give such students a preliminary view of the rich intellectual fields that lie before them, so that, on the one hand, all of their work shall have a large measure of unity and coherence, and on the other hand, they will be able to decide early what particular subjects they may wish more thoroughly to explore. There has been full co-operation between the various experts engaged in this educational experiment, and its success has seemed to make advisable the production of this volume. The arrangement adopted is from the general to the specific. Beginning with I. Astronomy (by Forest Ray Moulton), II. The Origin and Early Stages of the Earth (by Rollin T. Chamberlin) leads naturally to III. Geological Processes and the Earth's History (by J. Harlen Bretz). Physics and Chemistry follow, represented by IV. Energy: Radiation and Atomic Structure (by Harvey Brace Lemon), and V. The Nature of Chemical Processes (by Julius Stieglitz). Biology commences with VI. The Nature and Origin of Life (by Horatio Hackett Newton), followed by VII. a chapter on The Bacteria (by Edwin O. Jordan), after which Merle C. Coulter and Henry Chandler Cowles deal respectively with VIII. Evolution of the Plant Kingdom, and IX. Interaction between Plants and their Environment. General Zoology finds expression in X. The Evolution of the Invertebrates (by W. C. Allee) and XI. The Evolution of the Vertebrates (by Alfred S. Romer), after which XII. The Coming of Man is described by Fay-Cooper Cole. H. H. Newman (the Editor of the series) then discusses XIII. The Factors of Organic Evolution, and Elliot R. Downing XIV. Human Inheritance. XV. Man from the point of view of his Development and Structure (by George W. Bartelmez), XVI. The Dynamics of Living Processes (by Anton J. Carlson), and XVII. Mind in Evolution (by Charles H. Judd), conclude the series. A few selected references are given at the end of each chapter, and a somewhat limited glossary precedes the index. The volume is well printed and the illustrations, which in part are photographic reproduction in the form of plates and in part simple line text-figures, are clear and helpful. The book should have a wide appeal to readers who seek a general knowledge of the world about them, and take an interest in the history and future of their own species. XXIV. Proceedings of Learned Societies GEOLOGICAL SOCIETY. [Continued from vol. iv. p. 256]. November 2nd, 1927.--Dr. F. A. Bather, M.A., F.R.S., THE following communication was read:-- 'The Stratigraphical Distribution of the Cornbrash: I.—The South-Western Area.' By James Archibald Douglas, M.A., D.Sc., Sec.G.S., and William Joscelyn Arkell, B.A., B.Sc., F.G.S. This paper gives an account of the stratigraphical distribution of the Cornbrash in South-Western England, from Oxford to the south coast near Weymouth. The evidence obtained by the Authors is used in a critical discussion of the eleven brachiopod zones proposed by Mr. S. S. Buckman in a recent communication to the Society, and as a result a number of amendments in the list are suggested. A general description is given of the faunal sequence, and the distribution of the zones throughout the area is indicated by a detailed account of many type-exposures. The new records obtained are added to Mr. Buckman's faunal range-diagram, and in this way it is shown that many of his conclusions regarding penecontemporaneous erosion and nonsequences in the Cornbrash, as expressed in his clinal diagram, have been based on insufficient data, and that lack of geological information, in many instances due merely to exposure-failure or faulty collection, has been used as evidence of stratal failure. The introduction of new stratigraphical terms is criticized, and a twofold rather than a threefold subdivision of the Cornbrash is advocated. The more important brachiopods and lamellibranchs are discussed in paleontological notes, and several new species are described and figured. [The Editors do not hold themselves responsible for the views expressed by their correspondents.] |