November 5, 1908] 



A^A TURE 



25 



inductive resistance R', vvliicli is adjusted until the ampli- 

 tude of oscillation is the same in both cases. Then the 

 self-inductance is R' divided by the frequency-constant of 

 the alternator. The strength of the current- involved in 

 this measurement is known by imitating the deflection 

 with a known steady current. 



Prof. A. M. Worthington then showed a remarkable 

 series of instantaneous photographs exhibiting a new 

 feature in the splash of a rough sphere. This new feature 

 appears when the height of fall is increased beyond a 

 certain critical value. Below the critical height the splash 

 is characterised by an upward jet thrown high into the 

 air. It is now found that when the critical height is 

 passed the long cylindrical column of air which follows 

 the sphere in its descent through the liquid is pierced by 

 a central downward jet directed from above along the 

 axis of the air column. This is due to the permanent 

 closing, at an early stage, of the mouth of the air column 

 bv a film of the liquid, and to the subsequent reduction of 

 the pressure of the confined air through the piston-like 

 action of the sphere when its momentum is large enough. 

 The morning's proceedings concluded with a paper by 

 Prof. F. T. Trouton on the analogy between adsorption 

 from solutions and aqueous condensation on surfaces. 

 When cellulose is inserted into the solution of a dye 

 adsorption takes place, the amount of which depends upon 

 the concentration and the temperature, but the amount 

 can be kept at any particular value by simultaneously 

 varying both. When such corresponding values of con- 

 centration and temperature are plotted against one another 

 the curves are similar to one another, and, further, they 

 are similar to the ordinary saturation curve for the solute 

 in question. This result is analogous to the law of the 

 temperature isoneres for water vapour when we sub- 

 stitute osmotic pressure for concentration and the satura- 

 tion curve of the solution for the boiling-point curve, viz. 

 that at different temperatures the pressure ordinate of a 

 given isonere is a constant fraction of the corresponding 

 ordinate of the boiling-point curve. Thermodynamical 

 considerations were given in favour of both results. 



On resuming the sitting in the afternoon a paper by 

 Dr. J. A. Harker and Mr. F. P. Sexton was read (by the 

 former), on the effect of pressure on the boiling point of 

 sulphur. The results are closely represented by the 

 formula 



T = T,+O'ogo4 (/- 760) -0-0000519 (/-760)-, 



where T is the temperature of the vapour on the air-scale 

 at the pressure p in mm., and T, is the normal boiling 

 point. This gives a result much greater than the value 

 0082 mm. per degree which is usually employed, and 

 which is based on Regnault's observations. 



Dr. Glazebrook then communicated a paper on the 

 photometric standard of the National Physical Laboratory. 

 Wet- and dry-bulb thermometers are found to give results 

 20 per cent, higher for the humidity of the air than hygro- 

 meters of the Assman pattern, which are used at the 

 Reichsanstalt. The former were used at the National 

 Physical Laboratory in connection with the effect of 

 humidity on the pentane lamp. It is proposed to change 

 the standard humidity from lo to 8 litres per cubic metre, 

 and thereby maintain the light value unchanged. 



A paper by Mr. John Brown, on a dry Daniell pile, was 

 taken as read in the absence of the author. 



Meanwhile, the department of cosmical physics had been 

 meeting, the first paper being by Sir John Moore, on the 

 question. Is our climate changing? The object of the 

 paper was to test the accuracy of the popular opinion 

 that there is a progressive postponement of season, an 

 opinion strengthened by occasional abnormal weather con- 

 ditions, such as the snow and frost at the end of April, 

 igoS, and the summer heat at the beginning of September, 

 1906. From an examination of old records and of the 

 long series of observations made at Greenwich, the con- 

 clusion was drawn that no appreciable change has taken 

 place in our climate during the past six centuries. 



Dr. Shaw pointed out as instances of progressive changes 

 bearing on this question the gradual receding of glaciers 

 and of the Antarctic ice barrier, which had lost thirty 

 miles in tftn years. 



Coinmander Campbell Hepworth, C.B., of the Meteor- 

 NO. 2036, VOL. 70] 



ological Office, read a paper on the changes in the 

 temperature of the North .'\tlantic and the strength of the 

 trade winds. The N.E. trade wind is strongest in April 

 (135 miles per hour) and weakest in September (7-4 miles 

 per hour). The S.E. trade wind is strongest in February 

 (15-5 miles per hour) and weakest in May (13-7 miles per 

 hour). 



The surface temperature was lowest in March and 

 highest in August. 



There appears to be a relation between the departures 

 from mean velocity in the trade winds in one year and 

 the departures from mean temperature in the surface 

 waters in the succeeding year. 



A paper by Mr. F. J. M. Stratton, on the constants of 

 the lunar libration, described how a re-investigation of the 

 heliometer observations of Mosting A made by Schluter at 

 Konigsberg in the years 184 1-3 has been undertaken in 

 the hope of reconciling the conflicting sets of constants 

 given by Drs. Franz and Hayn. 



Mr. W. Makower, .Miss Margaret White, and Mr. E. 

 Marsden contributed the results of observations on the 

 electrical state of the upper atmosphere. The current down 

 a kite wire when the kite is at an altitude of 1500 metres is 

 of the order of axio--" amperes. It increases with the 

 height more quickly than according to the linear law, and 

 varies in a more or less regular way with the wind velocity. 



On Tuesday, September 8, the section was also divided 

 into three parts. In the mathematical department two 

 papers were contributed by Prof. A. W. Conway. In the 

 first— application of quaternions to problems of physical 

 optics — Prof. Conway showed how the analytical treat- 

 ment of such problems becomes both simpler and more 

 elegant when they are expressed in quaternion notation. 

 As examples he worked out the problem of reflection and 

 refraction at a plane surface, showing how to obtain the 

 ratio of the intensities ; and also that of the propagation 

 of light through a rotationally active medium such as a 

 sugar solution. 



Prof. Conway's second paper dealt with the distribution 

 of electricitv jn'a moving sphere. The sphere was assumed 

 of invariable form, and its velocity less than the velocity 

 of light. In the discussion which followed. Prof. Conway 

 mentioned that Mr. Varley had recently found that a 

 point of inflection in the curve of mass to velocity _was 

 indicated by experiment, and no theory could be entirely 

 satisfactorv which did not show such an effect. 



Major P. A. MacMahon read a paper on a problem 

 known as that of the " Scrutin de Ballotage." This 

 problem relates to the probability that when two candi- 

 dates are up for election, the candidate finally successful 

 shall be throughout at the top of the poll. Major 

 MacMahon has generalised this by considering an election 

 where there are any number of candidates, and has found 

 the probability that at any time during the election the 

 candidates shall be in the same order as they are finally. 



Prof. R. W. Genese followed with a paper on the 

 analysis of projection. He showed that if the vanishing 

 lines of two figures in space perspective be taken as axes 

 of y, Y respectively, and the lines where the planes of the 

 two figures are met by a plane through the vertex of 

 projection perpendicular to both as axes of x, X respec- 

 tively, then the coordinates are connected by the relations 



^ = ■"' = 1 



V z x' 



s, Z being constants, which may be taken as unity, and 

 the curve y = /(.v) in one plane transforms into the curve 

 y = xf(ifx) in the other. 



Mr. H. Bateman then explained a method of obtaining 

 solutions of problems in geometrical optics by conformal 

 transformations in space of four dimensions. He showed 

 that for such transformations (of which inversion is an 

 important particular case) the equations 



and 



dji' fh'^ dz- dw- 

 are invariant, and consequently from any one solution of 

 such equations a new solution can be at once deduced. 



