February 24, 1910J 



NA TURE 



.507 



of the benches have been kept as clear as possible, carry- 

 ing only Bunsen burners and two movable trays, each of 

 which can hold ten reagent bottles. These trays, with the 

 bottles they contain, fit into cavities in the sides of the 

 benches, so that the tops can be cleared in a moment when 

 required for practical geographical work. The reagent 

 bottles are double-labelled, so that they can be used by 

 two pupils working opposite each other. Attached to the 

 front of the demonstrator's bench is a shelf, which hangs 

 vertically when not in use. This has been designed for 

 the purpose of holding the apparatus required for the 

 lesson, and two girls from each bench come to this dis- 

 pensing shelf and take from it all they require for their 

 experimental work. The laboratories are both supplied 

 with many light trays of varying sizes, each capable of 

 holding a dozen beakers, flasks, burettes or pipettes, &c. 

 These trays fit into the bench cupboards. Neither cup- 

 boards nor drawers have been set apart, as is usual, for 

 the individual use of the pupils except in the case of the 

 more advanced students, as experience has shown that 

 they are apt to become receptacles for burnt matches, 

 corks, soiled filter papers, &c. Placed in each stool recess 

 is a shelf which holds a trough, test-tube rack, tripod, and 

 retort-stand. The same principle has been observed in the 

 fittings of the physical and botanical laboratory. The 

 fume cupboards, of which there are four — two in the 

 chemical laboratory, one in the physical laboratory, and 

 one between the lecture theatre and the preparation room — 

 are all supplied with both gas and water. The building 

 is fitted throughout with electric light. In the lecture 

 theatre there is an electric lantern, and a part of the 

 cream-coloured wall acts as the screen, and allows a 

 picture lo feet square. This room is fitted with dark 

 blinds, and ventilated, when these are in use, by means of 

 an electric fan. In the conservatory are benches at which 

 the pupils work when fitting up apparatus for botanical 

 experiments. The usual precautions have been taken 

 against accident by fire, and Minimax fire extinguishers 

 stand in prominent positions. 



SOCIETIES AND ACADEMIES. 



London. 

 Royal Society, January 13. — D. P. Phillips: Re-com- 

 bination of ions at different temperatures. The ionisa- 

 tion was produced in a layer of air of uniform thickness 

 by means of a single discharge through a Rontgen bulb. 

 The layer of air ionised was situated midway between two 

 parallel electrodes, and was separated from each by a layer 

 of un-ionised air. The quantity received by each electrode 

 depends upon the field established between them, and from 

 the variation of the quantity with the field the coefficient 

 of re-combination can be calculated. By placing the pair 

 of electrodes in a double-walled jacket the temperature was 

 varied, and the coefficient of re-combination found at 

 dififerent temperatures. The values which were found 

 are : — 



Temp. Centigrade ... 16' 100° 155° 176° 273° 

 CoefT. of Re-combinalion 100 0'50 0^399 036 o'i7S 



The value at the temperature of the room, i.e. at 16°, was 

 taken as unity, and the other values were compared with 

 this. The object of having the layer of ionised air 

 separated from the electrodes by un-ionised air was to 

 decrease the number of ions reaching the electrodes by 

 diffusion, and so causing an apparent increase in the re- 

 combination. With this arrangement the effect of diffusion 

 would be to decrease the apparent re-combination. In 

 order to test the magnitude of the error introduced by 

 diffusion, the thickness of the ionised layer of air was 

 altered, and the coefficient of re-combination determined 

 for each thickness. At each temperature it was found that 

 the coefficient of re-combination apparently falls off when 

 the thickness of the layer is reduced below a certain value. 

 Thus it was shown tliat in this experiment the diffusion 

 was negligible up to 176° C, but that at 273° C. it prob- 

 ably caused a serious reduction in the apparent value of 

 the re-combination. 



NO. 2104, VOL. 82] 



Sir Edward Thorpe and .\. G. Francis : The atomic 

 weight of strontium. The principle of the methods 

 employed consisted in determining the ratios of the weights 

 of strontium bromide and chloride and of pure silver and 

 uf the silver halides respectively. The strontium salts, 

 SrBr, and SrCU, purified by fractional crystallisation and 

 precipitation, were fused in a stream of dry halogen acid 

 and allowed to solidify in dry nitrogen. While the halides 

 were still warm the nitrogen was replaced by dry air and 

 the salts transferred to the weighing flasks. The fused 

 salts were ice-like in appearance, and yielded perfectly 

 clear neutral solutions in water. The silver needed to 

 precipitate completely the halogen was dissolved in _ a 

 specially devised burette, so contrived that the solution 

 could be delivered without loss to the strontium solution. 

 After eighteen hours the slight excess silver left in solution 

 was titrated with a solution of strontium halide of known 

 strength. Finallv, the silver halide was dried, fused, and 

 weighed. The apparatus was so devised that these opera- 

 tions could be done without removing the silver salts from 

 the vessel in which it was formed. As an independent 

 check, the ratios of SrBr, and SrCl, to SrSO, were also 

 determined by converting the strontium halides into 

 strontium sulphate by direct treatment with sulphuric acid. 

 The possible sources of error are discussed, and all known 

 corrections were applied. In all, six series of observations 

 were made. The mean results are as follows : — • 



Series A. 2Ag tSrBr., (6 exp's.) 87-645 + 00037 



,, B. 2AgBr: SrBr,, (5 expts.) 87-65^±00045 



,, C. 2Ag:SrCl., (6 expt«.) 87-642 + 0-OOI7 



„ D. 2A£;C1 :SrCl., (5 expts.) 87-645±ooo20 



„ E. S.Br.,:SrSOJ (3 expts.) 87629 + 0021 



,, F. SrCl, :PrS04(4 expis.) 87-6()i +00078 



Mean of A, B, C, D 87-646±oooi6 



E, F 87-645 + 0-OI07 



'„ A, B, C, D, E, F 87-646+00029 



The authors adopt 87-65 as the definite value for the atomic 

 weight of strontium— a number only 0-03 in excess of 

 Richards 's final value as given in the last report of the 

 International Committee on Atomic Weights. 



February 17.— Sir Archibald Geikie, K.C.B., president, 

 in the chair.— E. Marsden : Phosphorescence produced by 

 a- and y3-rays.— Prof. E. Rutherford : Theory of the 

 luminosity pi^oduced in certain substances by a-rays. — Dr. 

 H. Geiger : The scattering of the a-particles by matter. 

 In a previous note on the same subject experiments have 

 been described which gave direct evidence of the scattering 

 of the a-particles in passing through matter. These experi- 

 ments have been continued with the object of determining 

 quantitatively the amount of scattering under various con- 

 ditions. In particular the influence of the thickness and 

 nature of the scattering material and of the velocity of the 

 a-particles has been studied in detail. With the exception 

 of a few modifications, the experimental arrangement was 

 the same as that employed in the preliminary experiments. 

 A strong source of homogeneous a-radiation was placed at 

 one end of a long tube, and the a-particles, after passing 

 through a narrow circular opening, fell upon a zinc sulphide 

 screen sealed to the other end of the tube. When the 

 pressure inside the tube was very low the scintillations 

 produced by the impact of the a-particles on the screen were 

 confined to a %-ery small area. When, however, the 

 a-particles were intercepted by a thin sheet of metal, the 

 scintillations were spread out over a much greater area, 

 this being due to the scattering of the a-particles when 

 passing through the metal sheet. The distribution of the 

 scintillations over the screen was determined by counting 

 them at different parts of the screen. From the distribu- 

 tion curve the most probable angle through which the 

 a-partlcles were turned in passing through the metal sheet 

 under investigation could be found. In all experiments the 

 scattering was measured by this angle, and the following 

 results were obtained : — (i) The most probable angle of 

 scattering increases for small thicknesses approximately 

 proportional to the square root of the thickness of matter 

 traversed by the o-partlcle. For greater thicknesses the 

 scattering angle increases more rapidly. (2) The most 

 probable angle of scattering is proportional to the atomic 



