October 2 i, 1909] 



NA TURE 



497 



Konen, of Munster, and his principal results are sum- 

 marised in No. 17 of the Verhandlungcn der deutschen 

 physikalischen Gesellschaft. A comparison has been made 

 between the spectra of more than twenty metal electrodes 

 in air and in water, the spari< being produced by an 

 induction coil of 40 cm., having a capacity of 002 micro- 

 farad in parallel with its spark gap. The spectra in air 

 and in water were photographed on the same plate, a 

 concave grating of radius 180 cm. being used. The spectra 

 in water show no air lines, but a large proportion of them 

 show a more or less extended continuous spectrum with 

 the lines of water vapour reversed. Some lines are un- 

 changed, while others are broadened on one or on both sides, 

 lines belonging to the same series being modified in the 

 same sense. No trace of the spectra of salts in solution 

 in the water has in any instance been detected. 



Part iii., vol. xvi., of the Proceedings of the Cotteswold 

 Naturalists' Field Club maintains the high standard of 

 style and illustration which distinguishes this publication. 



streams, formed in the drainage area before any definite 

 line of principal stream has been settled." .Accounts are 

 given of places and objects of scientific interest visited 

 during e.xcursions, and among these is a record of an 

 excursion to Shepton Mallet and Vallis Vale, which is 

 well known for its romantic scenery and geological un- 

 conformities. The report contains the accompanying illus- 

 tration, reproduced by permission of the council of the 

 Geological Society, of the classic section showing Oolite 

 resting upon Carboniferous Limestone, which was pictured 

 by Sir H. de la Beche. Short papers by Messrs. L. 

 Richardson and C. Upton, and a report by the Rev. H. J^ 

 Riddelsdell, on the progress made in connection with the 

 flora of Gloucestershire, complete the issue. 



Le Radium for August contains a very useful table of 

 the principal minerals containing uranium and thorium, 

 prepared by M. B. Szilard, of Madame Curie's laboratory. 

 It occupies seven pages of the periodical, and gives the 

 name, the composition, the percentages of uranium and of 



View of Ouariv. showing ihe Inferior Oolite resting unconform.-ibly upon the Carboniferous Limestone in Vallis Vale. Photograph by Prof. 



.S. H. Reynolds. 



A large part is concerned with the record of excursions 

 chiefly to places of geological or antiquarian interest, and 

 is illustrated by no fewer than nine plates of photographic 

 reproductions. Of the papers published, the longest is on 

 " The Lower Severn : Valley, River, and Estuary from the 

 Warwickshire to the Bristol Avon," by Mr. T. S. Ellis. 

 The principal aim of the author is to show that the 

 generally accepted views of river development associated 

 with the name of Prof. W. M. Davis are not applicable 

 to the district with which he deals, and by inference to 

 other districts also. His position is indicated by the follow- 

 ing sentence : — " As I believe, we cannot have a right 

 conception of the development of rivers unless we keep our 

 minds free from all idea of original lines of streams or of 

 any principal line as the initial condition. In my view, 

 a river system is evolved not 111(0, but from a network of 

 NO. 2086, VOL. 81] 



thorium, the localities in which the mineral is found, and 

 its crystalline form. In addition, a map of the world, in 

 which the localities are shownj is given. 



Mr. Harold Moore, of Woolwich Arsenal, read a paper 

 on the Brinell method of determining hardness at the Inter- 

 national Association for Testing Materials. We note from 

 an abstract in Engineering for October 8 that the author's- 

 results show that the thickness of the test specimen has 

 no influence on the result provided that the depth of the 

 impression made by the ball does not exceed one-seventlr 

 the thickness of the specimen. A safe rule to be adopted 

 is that the distance of the centre of the impression from 

 the edge of the specimen should not be less than 2-5 times 

 the diameter of the impression. For calculating the hard- 

 ness number, 30° should be chosen as the standard angle 

 of impression, this giving a diameter of impression equal 



