216 Scientific Intelligence. 



The chief experimental difficulty consisted in keeping the X-ray 

 bulb sufficiently " hard " when the anticathode was covered with 

 metals of low melting points, such as indium « 200° C). This 

 was overcome by causing a stream of cold water to play against 

 the back of the cathode and by running the Gaede mercury pump 

 continuously, charcoal in carbon dioxide and ether being used 

 simultaneously. A rock-salt crystal was employed in the usual 

 way to analyze the radiations. 



When the square roots of the frequencies are plotted as 

 ordinates against the associated atomic numbers as abscissae it is 

 found that the points lie close to two straight lines which cor- 

 respond respectively to the a and fi spectral lines. On the con- 

 trary, these right lines have not the same slopes as the lines 

 derived from Moseley's data. Consequently the relation 

 \v == fv (A r — l) 2 ] discovered by Moseley for the elements from 

 aluminium to zinc does not continue to hold good throughout the 

 series of the heavier elements. Furthermore, it should be 

 remarked that antimony fits in with the rest of the elements in- 

 vestigated by Maimer only when it is assigned the atomic num- 

 ber 52, whereas it must have the value 51 to agree with the 

 L-series. If this be true, an element may have one atomic num- 

 ber in the K-series and a different number in the L-series. In 

 the case of cerium no lines of the K-series were found, although 

 the bulb was extremely hard. Finally, the lines of silver were 

 studied in detail to find out if they are doublets such as have 

 been recorded by Bragg for rhodium and by Rawlinson for 

 nickel. The first negative showed two a-lines and also duplicity 

 in the direct image. This doubling was due to the fact that the 

 cathode rays were not focussed in only one point of the target. 

 "Consequently, the bulb was placed in such a position that the 

 X-rays used were emitted at a minimal angle with the surface of 

 the target, and now both the direct and the reflected lines were 

 simple." "Possibly former cases of double lines may be referred 

 to the same cause." — Phil. Mag., xxviii, p. 787, December, 1914. 



h. s. u. 



7. The Bifilar Property of Twisted Strips. — When a strip of 

 phosphor-bronze, carrying an inertia bar, is twisted and then 

 allowed to oscillate, its period of vibration is found to depend 

 upon the load. As is well known, the period of a bifilar suspen- 

 sion depends upon the load and hence it has been suggested that 

 a bifilar action is present in strips and accounts, in part at least, 

 for their behavior. Since the results obtained by Campbell and 

 by Pealing were not altogether concordant, the question has been 

 taken up both theoretically and experimentally by J. C. Buckley. 



• ■ , , „ 16bd*?i 2Tb' . „ , . ' ' 



The formula C = 1- — = — is first derived, where 2o = 



31 31 



horizontal breadth of strip, 2d = thickness of strip, I = length, 

 n = constant of torsional rigidity, C = restoring couple per unit 

 angular displacement, and 7'z= tension per cm. The first term of 

 the right-hand side of the equation corresponds to the twisting of 



