NOW READY. Cloth, 316; Paper covers, 216. (Postage, 4d. extra).

THE WHEAT PROBLEM: Based on Remarks made in the Presidential Address

to the British Association at Bristol in 1898.

REVISED WITH AN ANSWER TO VARIOUS CRITICS

By SIR WILLIAM CROOKES, F.R.S.

SECOND EDITION.

VITH PREFACE AND ADDITIONAL CHAPTER, BRINGING THE STATISTICAL INFORMATION UP TO DATE.

With Two Chapters on the Future Wheat Supply of the United States, by MR. C. WOOD DAVIS, of Peotone, Kansas, and the HON. JOHN HYDE, Chief Statistician to the Department of Agriculture, Washington.

OPINIONS of the PRESS.

"In his bulky volume Sir William reproduces the gist of the sensational Bristol Address, and supplements it with carefully prepared answers to his chief critics and confirmatory chapters on the future wheat supply of the United States."-Morning Post.

"The fuller examination of the problem as here conducted shows that Sir William Crookes did not speak unadvisedly with his lips."-Yorkshire Post.

"The problem is one of importance, and Sir William Crookes presents it to us fortified by the opinions of two American experts."-Manchester Guardian.

"Sir William Crookes's statistics seem to make good his alarmist statement."-British Weekly.

"In the present volume Sir William Crookes replies vigorously to his critics."-Liverpool Daily Post.

"The book is a useful one to all interested in the production of wheat both from the commercial and scientific points of view."-Knowledge.

"It is a vital question, and considering the cheap issue of the volume all interested in the feeding of the millions ought to get it and read it carefully."-Crieff Journal.

"If these somewhat gloomy prognostications result in drawing the attention of chemists more seriously to what has hitherto been only an interesting laboratory problem, Sir William Crookes will have conferred an incalculable benefit on the race."-Western Morning News.

"Sir William discusses at length the criticisms passed upon his address, and he appends valuable papers supporting his arguments on the future wheat supply of the United States."-Globe.

"The student of economic science and sociology will find this volume full of interesting material. The entire subject is of the profoundest interest, and an excellent pur pose has been served by the publication of these papers in a single volume."-The Eagle (Brooklyn, N.Y.)."

CHEMICAL News, Jan. 8, 1909

}

THE CHEMICAL NEWS.

VOL XCIX., No. 2563.

MEASUREMENT OF ROTATORY DISPERSIVE POWER IN THE VISIBLE AND ULTRA-VIOLET REGIONS OF THE SPECTRUM.*

By T. MARTIN LOWRY, D.Sc, Lecturer on Physical Chemistry and Instructor in Crystallography at the Central Technical College.

THE following is a brief preliminary account of improvements effected in the method of determining rotatory dispersive power which have made it possible to observe accurately not only in the bright regions of the visible spectrum, but throughout the scale from the region of the lithium red line into that commanded by the photographic plate.

Two methods have generally been used for the purpose, namely, (1) Broch's method, in which a spectroscope is arranged in series with the polarimeter and a narrow strip of a continuous spectrum is picked out for observation-a method which is much improved by using a constantdeviation spectroscope in place of one of the variabledeviation type (F. Twyman, Phil. Mag., 1907, xiii., 481), and (2) Landolt's method, in which a white light is reduced by means of filters to approximate homogeneity in the red, green, light blue, or dark blue parts of the spectrum. Neither method fulfils the fundamental condition that the field of the polarimeter shall be uniformly lighted with monochromatic light-many of the measurements that have been made, therefore, possess only a qualitative value. A much better method is due to the late Sir William Perkin, who introduced the use of a spectroscopeeyepiece as a means of purifying the sodium light, and used it on a limited scale for measuring rotatory dispersive power in the red (lithium), yellow (sodium), and green (thallium) parts of the spectrum.

The method now described resembles Perkin's method in its essential feature, namely, the use of monochromatic or multichromatic light (spectroscopically purified) in place of a band from a continuous spectrum. It has the advantage that it renders available for polarimetric measurements, in addition to the flame spectra, the large series of intense line spectra produced by the metallic arcs, which, with the one exception of the enclosed mercury arc (Disch, Ann. Phys., 1903, [4], xii., 1155), do not appear to have been used previously for this purpose. Up to the present, measurements have been made with 26 lines ranging from wave-length 6708 to 4359; beyond these limits the visual intensity of the light is so small that polarimetric observations become very difficult, but in the intermediate part of the spectrum the list might, without difficulty, be extended considerably. The mercury lines referred to in the table were produced by a Bastian lamp; the copper and zinc spectra were produced by means of copper or brass electrodes rotating in opposite directions, as recommended by Baly in the case of the iron arc; the cadmium spectrum was obtained by means of rotating copper or silver electrodes coated with the metal.

In order to utilise the arc spectra for polarimetric observations, a parallel beam from the arc is thrown on to the widely-opened slit of a constant deviation spectroscope. An achromatic lens of 22" focus is substituted for the observing telescope of the spectroscope, and is used to cast a magnified image of the slit on to the polarising prisms which produce the horizontally-divided triple field of the polarimeter. By turning the constant-deviation prism to a suitable position, the field can be illuminated from top to bottom by a brilliant band of pure monochromatic light, the maximum width of the band being * A Paper read before the Royal Society, November 19, 1908.

13

determined by the openness of the spectrum and, in a very important manner, by the efficiency of the dispersive

system. By using a C.D. prism of high density in conjunction with the long-focus lens, it is possible to read separately the two lines 5790 and 5769 which constitute the yellow mercury doublet, although these differ in wavelength by only 21 Angstrom units; the yellow doublet and green triplet in the copper spectrum can be read with ease and accuracy as broad bands each occupying a width rather greater than one-third of the 8 mm. aperture of the triple field; the chief lines in the mercury and cadmium spectra can be made to cover practically the whole field without overlapping. In order to eliminate stray light, which would give rise to serious errors in the red and violet readings, a Perkin spectroscope eye-piece is used. In measuring rotatory dispersive power in the ultraviolet, a parallel beam of light from an arc formed between a carbon and a magnetite electrode is cast directly by a quartz condenser on to the triple field of the polariser, Foucault prisms being substituted for the Nichol prisms to ensure transmission. A quartz-calcite lens of 13" focus is substituted for the analyser-telescope; this casts a diminished image of the triple field on the slit of an ultraviolet spectroscope. The spectrum thus produced is photographed in the ordinary way by means of a camera provided with a quartz calcite lens of 22" focus; the division produced by the triple field can be seen clearly, and it is easy to pick out and identify on the negative the line which is of equal intensity in its three sections. By taking photographs with the analyser in different positions it is possible to determine the rotatory power of a substance throughout the transmitted spectrum.

The results obtained with water, carbon bisulphide, and other liquids in a magnetic field will be dealt with in a later communication; the method is one which is likely to be of special value as affording a means of determining the below the results are reliable within two or three units in homogeneity of apparently simple liquids. In the table the last figure.

concordant within a few hundredths of a degree; the The polarimeter readings shown in the table were usually absolute values may differ by as much as 1° in different specimens of the ester, but the relative values, a D, are probably reliable within one or two units in the last figure.

Table of Wave-lengths used in the Measurement of Rotatory Dispersive Power, together with the Rotations produced by 100 mm. of Methylic Camphocarboxylate at 20° C.

THE SPECTRUM OF SCANDIUM AND ITS RELATION TO SOLAR SPECTRA.*

By A. FOWLER, A R.C.S, F.R.A.S., Assistant Professor of Physics, Imperial College of Science and Technology, South Kensington.

THE greater part of this investigation of the spectrum of scandium under different experimental conditions has been | based on purified scandia, generously placed at the author's disposal by Sir William Crookes. The principal results are as follows:

1. The arc spectrum of scandium consists of two distinct sets of lines, which behave very differently in solar spectra. Each set includes both strong and faint lines.

2. Lines belonging to one set correspond with the enhanced lines of other elements, notwithstanding that they appear strongly in the ordinary arc spectrum

(a) These lines are very feeble or missing from the arcflame spectrum, and are strengthened in passing to the arc, the arc in hydrogen, or the spark.

(b) They occur as relatively strong lines in the Fraunhofer spectrum.

(c) They are weakened in the sun-spot spectrum. (d) They occur as high-level lines in the chromosphere. 3. The remaining lines show a great contrast when compared with the first group

(a) They are relatively strong lines in the arc-flame. (b) They are very feebly represented in the Fraunhofer spectrum.

spectrum.

(c) The stronger lines are prominent in the sun-spot (d) They have not been recorded in the spectrum of the chromosphere.

4. The special development of the enhanced lines in the Fraunhofer spectrum, together with their presence in the upper chromosphere, indicates that the greater part of the scandium absorption in the solar spectrum originates at a higher level than that at which the greater part of the iron absorption is produced.

5. The discussion of scandium lines indicates that while in the case of some elements solar identifications are to be based chiefly on arc lines, in others it is the enhanced lines which may be expected to show the most important coincidences.

6. The flutings which occur in the arc and arc-flame do not appear when the arc is passed in an atmosphere of As suggested by Thalén, they are probably hydrogen.

due to oxide of scandium.

Tables are given which show the lines of the arc spectrum from 3930 to 6580, the positions of the oxide flutings, and comparisons of the principal lines of the two classes with the sun, sun-spots, and chromosphere.

Decomposition of Diazo Solutions.-A. Hantzsch and K. J. Thompson.-After diazonium chlorides have been left for a long time in the desiccator, or dry air has been passed over them, they give the maximum value of the velocity of decomposition in aqueous solution. Freshly prepared solutions decompose more slowly and seem to contain some unknown substance which hinders decomposition. Dilute bromide solutions decompose at the same rate as the chloride. The velocity of decomposition in creases very slowly with the concentration, that of the bromide solution rather more quickly than the chloride. Diazoiodide solutions even when very dilute decompose much quicker, and nitrobenzene diazonium solutions also decompose rather more quickly in very concentrated solutions; acids do not affect the decomposition of the last named compounds.-Berichte, xli., No. 14.

* Abstract of a Paper read before the Royal Society, June 25, 1908.

THE CORROSION OF IRON.*

By ALLERTON S. CUSHMAN, Assistant Director, Office of Public Roads, Dept. of Agriculture, U.S.A. (Continued from p. 11).

The Peroxide Theory.

DUNSTAN, Jowett, and Goulding based their peroxide theory of the rusting of iron on the general well-known theory of oxidation advanced by Traube (Ber., xviii., 1881). Thus the chemical reactions concerned in the formation of iron rust should be written.—

Fe + O2 + H2O = FeO+ H2O2,

2FeO + H2O2 = Fe2O2(OH)2 = Fe2O3, H2O.

The excess of hydrogen peroxide immediately reacts with the iron, forming a further quantity of rust :Fe+ H2O2 = FeO + H2O,

The

2FeO + H2O2 = Fe2O2(OH)2 = Fe2O3, H2O. Some of the evidence brought forward to substantiate this theory has already been referred to in a previous paragraph. It appears to derive some confirmation from the fact that delicate tests for hydrogen peroxide have been obtained during the slow oxidation of zinc and some other metals. On the other hand, in the case of iron these same delicate tests obstinately refuse to reveal even its transitory presence during the ordinary process of rusting. theory has been criticised by Divers (Proc. Chem. Soc., 1905, xxi., 251), Moody (Fourn. Chem. Soc., 1906, lxxxix. -cx., 720), and Cribb (Analyst, 1905, xxx., 225)—the first-named having pointed out that it is not tenable to argue that, because such substances as chromic acid and alkalis gradually destroy hydrogen peroxide, they must prevent its formation. For instance, ferrous sulphate is oxidised by free chlorine, but it does not prevent manganese dioxide and hydrochloric acid from reacting when brought together in its presence. Moreover, if the formation of hydrogen peroxide was a necessary stage in the rusting of iron, and this is inhibited by certain substances which destroy hydrogen peroxide, why is not the inhibition extended to strong reducing agents generally? The theory is an interesting and suggestive one, but in the author's opinion is not supported by the facts.

The Electrolytic Theory.

From the standpoint of the modern theory of solutions, all reactions which take place in the wet way are attended with certain re-adjustments of the electrical states of the that before iron can oxidise in the wet way it must first reacting ions. The electrolytic theory of rusting assumes pass into solution as a ferrous ion. The subject has been interestingly treated by Whitney (loc. cit., p. 10), who discussed it from the standpoint of Nernst's conception of the source of electro-motive force between a metal and a solution. When a strip of metallic iron is placed in a solution of copper sulphate, iron passes into solution and copper is deposited, this change being of course accompanied by a transfer of electrical charge from the ions of copper to those of iron. Hydrogen acts as a metal, and is electrolytically classed with copper in relation to iron. If therefore we immerse a strip of iron in a solution containing hydrogen ions, an exactly similar reaction will take place, iron will go into solution, and hydrogen will pass from the electrically charged or ionic to the atomic or gaseous condition. In such a system the solution of the iron, and therefore its subsequent oxidation, must be accompanied by a "precipitation or setting free of hydrogen. It is very well known that solutions of ferrous salts as well as freshly precipitated ferrous hydroxide are rapidly oxidised by the free oxygen of the air to the ferric condition, so that if the electrolytic theory can account for the original solution of the iron the explanation of rusting becomes an exceedingly simple one.

Bulletin No. 30, U.S. Department of Agriculture, Office of Public Roads.