cyclohexanols derived from each cresol are found to possess widely differing viscosities and the relation between the magnitude or the viscosities of the complete series of six inactive methlycyclohexanols and their acetic esters, is found to be in agreement with the views of Thole.

It has been found possible to effect the geometrical interconversion of the cis- and trans-methlycyclohexanols through the medium of the p-toluenesulphonic esters. Partly by this method and partly by the oxidation of the alcohols to methylcyclomexanols have been determined. Two distinct types of relationship are observed.

A study is also made of the effect of experimental conditions on the proportions of the two geometrical isomerides formed by various methods of reduction.

To receive and discuss the following papers :

66

Opal Glass." By J. W. Ryde. A communication from the Research Laboratories of the General Electric Co., Ltd. Wembley.

"The Electrical Conductivity of Glass at High Temperatures." By F. F. S. Bryson, M.A., BSc., F.Inst. P.

The Influence of Moisture on the Rate of Melting and on the Properties of SodaLime-Silica Glasses." By Edith M. Firth, B.Sc., F. W. Hodkin, B.Sc., Michael Parkin, M.Sc., and Prof. W. E. S. Turner. "The Physical Properties of Zinc Oxde By Containing Glasses." S. English, D.Sc., and Prof. W. E. S. Turner.

EIGHTH ANNUAL DINNER.

The eighth annual dinner of the Society will be held in the Hotel Cecil, Strand, London, W.C.2., on Tuesday, June 1, 1926, at 7.0 for 7.15 p.m. Tickets (excluding wines) 12s. 6d. each. Evening dress.

The guests from America are expected to number about 30 and it has been suggested as a happy arrangement that each visitor should have a host at the dinner.

THE INSTITUTION OF ELECTRICAL ENGINEERS.

Savoy Place, Victoria Embankment, W.C. Wireless Section Meeting.-Wednesday, June 2, 1926, at 6 p.m. (Light refreshments at 5.30 p.m.) "Low Frequency Intervalve Transformers." By P. W. Willans, M.A.

ROYAL SOCIETY OF ARTS. John Street, Adelphi, London, W.C.2. Indian Section. Monday, May 31, 1926, at 4.30 p.m. "The Military Record and Potentialities of the Persian Empire." By Lieut.-Col. Sir Arnold Talbot Wilson, K.C.I.E., C.S.I., C.M.G., D.S.O. Brig.General Sir Percy M. Sykes, K.C.I.E., C.B., C.M.G., will preside.

SOCIETY OF PUBLIC ANALYSTS AND OTHER ANALYTICAL CHEMISTS. An additional meeting of the Society will be held on Wednesday, June 2, at the Chemical Society's Rooms, Burlington House, Piccadilly, W.1., at 8 p.m., when Mr. Alfred Lucas, O.B.E., F.I.C., will read a paper on 66 Problems in Connection with Ancient Egyptian Materials."

SOCIETY OF GLASS TECHNOLOGY.

Ordinary General Meeting, June 1 and 2. First Session to be held in the University College, Gower Street, London, W.C.1., on Tuesday, June 1, 1926, at 2.30 p.m.

Second Session to be held in the Library of the Research Laboratories of the General Electric Co., Ltd., Wembley, on Wednesday, June 2, at 2.15 p.m.

CORRESPONDENCE.

THE ASSOCIATION OF NUMBERS IN CHEMISTRY.

The association of two or more atomic weights in numbers is frequently met with in the course of calculation of numerical relationships of the elements, and in other calculations which involve the use of atomic weights. No satisfactory method, however, appears to be known for systematically working out numerical associations of atomic weights. The writer considers his original method, described in this letter, as a perfectly satisfactory one for the purpose

(1) 64784/5897 × 3546 + 10238

49194/5897 × 238 =

(2) 64784/7992 × 3546 +

of working out by proportion numbers, which are formed by the association of atomic weights with atomic ordinals or with other whole number atomic weights.

I.

The number 64784 is the principal commencing number. It is actually the sum of 64546 + 00238. From it, numbers are derived by proportion with 3546/A, where A is an atomic weight. To the product, 00238 is added, and the number is reduced to five figures. Unity may be added in the first place, when it is necessary. (It is also usual to add unity beyond the fifth place, when accurate relationship is required, but this is not dealt with here.) The number obtained is multiplied by (a) 238/A; (b) 2377/A, where 238 is the three-figure approximate of 2377. The product is usually the association of atomic ordinal with the vertically adjacent atomic weight in the periodic system. The success of this process largely depends on the two original numbers. In this case they appear to be well chosen.

Examples

= 49194

=