402 Scientific Intelligence. 



possible sources of error, doubtless deeming them of negligible 

 importance. These errors may arise from (1) air streams within 

 the balance case, and (2) differences in the superficial areas of 

 the reaction vessels and the consequent possibility of a variation 

 in the relative masses of any aqueous films which may be formed 

 on the surfaces. Furthermore, evidence is lacking in Landolt's 

 memoir to show that he paid sufficiently close attention to effects 

 producible by very slight and varying differences in the tempera- 

 ture of the contents of anj r pair of his reaction vessels. Since 

 Manley considered these sources of error as extremely important 

 he accordingly introduced devices for eliminating and neutraliz- 

 ing the errors just indicated. 



A new and hitherto unsuspected secondary chemical reaction 

 occurring within the reaction vessels was discovered and partially 

 investigated. This reaction may be somewhat accelerated by 

 heat and greatly so by the radiations from a tantalum lamp. It 

 was shown that this phenomenon made it impossible to obtain 

 trustworthy results with solutions of silver nitrate and ferrous 

 sulphate. Decisive results could be obtained with solutions of 

 barium chloride and sodium sulphate because these salts react 

 quickly and practically perfectly. 



The limit of accuracy estimated by Landolt, under his working 

 conditions, was ±0-03 mgrm., while Manley claims ±0-006 

 mgrm., just five-fold better. Landolt found the mean apparent 

 change in mass during chemical reaction to be not greater than 

 1 in 1 X10 7 . In the most thoroughly investigated case, in which 

 the reacting bodies were barium chloride and sodium sulphate, 

 Manley found that the apparent change in mass did not exceed 

 1 in IX 10". — Proc. Roy. Soc, vol. lxxxvii, No. A 594, p. 202. 



h. s. v. 



9. On the Torque produced by a Beam of Light in Oblique 

 Refraction through a Glass Plate. — In the presidential address 

 to the Physical Society in 1905 Poynting developed the idea that 

 a beam of light must be regarded as containing a stream of 

 momentum, and he showed that this principle may be used to 

 solve with great ease the various cases in which a beam of light 

 is absorbed, reflected, or refracted at a surface. The particular 

 case in which a beam of light is refracted at a plane surface was 

 also analyzed by Poynting and the conclusion was drawn that 

 there will always be an outward pull along the normal, i. e. from 

 the more dense towards the less dense medium. Hence, when a 

 beam of light passes obliquely through a parallel plate there is a 

 normal pull outwards both at incidence and at emergence, and 

 these two pulls constitute a couple. Obviously, the forces arising 

 from the several reflections in any actual piece of apparatus must 

 be taken into account in making numerical calculations. Prelim- 

 inary quantitative experiments were performed by Poynting and 

 Barlow with the object of comparing the observed torque with 

 the value of the couple derived from the principle of luminiferous 

 momentum. However, the apparatus used did not yield very 

 satisfactory results. 



Since that time G. Barlow has devised and used an apparatus 



