56 



NA TURE 



[July 8, 19c 9 



the life and customs of other nations at a time when their 

 own opinions are forming, with a mitJtmum of incon- 

 venience to their academic worfc and the least possible 

 expense, with the view of broadening their conceptions 

 and rendering them of greater economic and social value. 

 Lord Strathcona has consented to become president for 

 the United Kingdom. The list of vice-presidents includes 

 the names of the Chancellors and Vice-Chancellors of 

 many British universities, the Prime Minister, the Lord 

 Chancellor, and other well-known men. A large and re- 

 presentative committee has also been appointed. Additional 

 objects of the movement are to increase the value and 

 efliciency of present university training by the provision 

 of certain travelling scholarships for practical observation 

 in other countries under suitable guidance. In addition to 

 academic qualifications, the selected candidate is what is 

 popularly known as an " all-round " man, the selection to 

 be along the lines of the Rhodes scholarships. It is 

 hoped to afford technical and industrial students facilities 

 to examine into questions of particular interest to them 

 in manufactures, &c., by observation in other countries 

 and by providing them with introductions to leaders in 

 industrial activity. It is proposed to establish two students' 

 travelling bureaux, one in New York and one in London; 

 to appoint an American secretary (resident in New York) 

 and a British secretary (resident in London), to afford 

 every facility to any graduate who wishes to visit the 

 United States, Canada, or the United Kingdom for the 

 purpose of obtaining an insight into the life of those 

 countries. It is hoped to provide twenty-eight travelling 

 scholarships, fourteen of these being available for universi- 

 ties in the United Kingdom, ten for universities in America, 

 and four for universities in Canada. The total cost of 

 the scheme, inclusive of the maintenance of two travelling 

 bureaux and the provision of twenty-eight scholarships per 

 annum, is estimated at 13,500?. for a period of three years, 

 equivalent to an annual expenditure of 4500Z., the relative 

 annual expenditure being estimated at 2400Z. in the United 

 Kingdom, 600/. in Canada, and 1500?. in the United States. 

 Promises of support may be sent to the hon. secretary, 

 Mr. Henry W. Crees, at the University Club, Birming- 

 ham, and it is hoped that all interested in promoting the 

 success of an educational scheme of far-reaching signifi- 

 cance in the English-speaking world will contribute 

 financially. 



SOCIETIES AND ACADEMIES. 

 London. 

 Royal Society, June 24. — Prof. J. Cossar Ewart, vice- 

 president, in the chair. — Pressure perpendicular to the 

 shear planes in finite pure shears, and on the lengthening 

 of loaded wires when twisted : J. H. Poynting. When 

 a solid is subjected to a finite pure shear the lines of 

 greatest elongation and contraction are not the diagonals 

 of the rhombus into which a square is sheared, but lines 

 making, respectively, +e/4 with the diagonals of the 

 square, where € is the angle of shear, and these lines 

 are at right angles to the order of e'. If we assume that 

 a pressure P is put on along the lines of greatest con- 

 traction, and a tension Q along the lines of greatest 

 elongation, we may put P = uf + pt', Q = !(e — />f-, where « 

 is the rigidity and p is a constant to the second order of €. 

 For equilibrium a pressure R = (lu + p)t- is required per- 

 pendicular to the shear planes. This is zero only if 

 />=— 2", a supposition for which there is no apparent 

 reason. To keep constant volume a stress may be needed 

 S = qf" perpendicular to the plane containing P and O, a 

 pressure if q is positive. Suppose that a wire is twisted 

 by a torque with axis along the axis of the wire. To keep 

 the volume constant at every point it would be necessary 

 to apply the system of forces R and S from outside. If 

 this system is not applied we may expect the wire to 

 change in length and diameter by amounts calculable in 

 terms of the elastic constants. The change in length 

 should be an increase dl = Sa-S-/l, where S is a function 

 of the constants, given in the paper, a is the radius, and 

 fl is the twist in length J. Such a lengthening has been 

 found to exist for piano-steel, copper, and brass wires 

 when loaded enough to straighten out kinks. For the 

 NO. 2071, VOL. 81] 



piano wires tested S was of the order i, and for the copper 

 and brass wires of the order 1-5. The lengthening of a 

 steel wire 0-97 mm. diameter and 23 metres long for a 

 twist of one turn in the length was about 0-0019 mm. 

 This lengthening on twisting should be taken into, account 

 in accurate determinations of the rigidity. — The wave 

 motion of a revolving shaft and a suggestion as to the 

 angular momentum in a beam of circularly polarised 

 light : J. H. Poynting. When a shaft of circular section 

 is revolving uniformly and is transmitting power uniformly, 

 a row of particles originally in a line parallel to the axis 

 will lie on a spiral of constant pitch, and the position of 

 the shaft at any instant may be described by the position 

 of this spiral. The motion of the spiral onwards may be 

 regarded as a kind of wave motion. Its velocity with a 

 given speed of revolution will only be the " natural " 

 velocity j/(ii;p) of twist waves along the shaft for a 

 certain torque on the shaft. For any other torque the 

 velociiv is " forced," and- forces from outside must be 

 applied to maintain it at every point where the twist is 

 changing. The group velocity of waves of this kind is 

 zero. Taking a uniformly revolving tube as a mechanical 

 model of a beam of circularly polarised light, and assuming 

 that the relation between torque and energy holding for 

 the model holds also for the beam of light, the angular 

 momentum delivered per second to unit area of an absorb- 

 ing surface upon which the light falls normally is PA/27r, 

 where P is the pressure of the light and X is its wave- 

 length. In light-pressure experiments P is detected by the 

 torque produced on a disc at the end of an arm about 

 I cm. The value is therefore about 100.000 times as great 

 as the torque or the same disc, due to the angular 

 momentum. If the angular momentum of circularly 

 polarised light only has this value, there does not appear 

 to be much prospect of detecting it at present. — The effect 

 of a magnetic field on the electrical conductivity of flame : 

 Prof. H. A. Wilson. This paper contains an account of 

 some experiments oi-i the change in the conductivity of a 

 Bunsen flame pioduced by a magnetic field. The current 

 through the flame was horizont,-il, and the magnetic field 

 was also horizontal, but perpendicular to the current. The 

 ratio of the potential gradient in the flame to the current 

 was taken as a measure of its resistance. The results 

 show that 5R/R = .AH'-|-BH, where H denotes the mag- 

 netic field. R the resistance, and A and B are constants. 

 The velocity of the negative ions can be calculated from 

 the term AH^, and the result is q6oo cm. /sec. for i volt 

 per cm., which agrees with Mr. E. Gold's results obtained 

 by an entirely different method. The term BH is pre- 

 sumably due to the upward motion of the flame gases, but 

 its value Is about fifty times greater than the value calcu- 

 lated from the Ionic theory. — Studies of the processes 

 operative in solutions : xi., the displacement of sails from 

 solution by various precipltants : Prof. H. E. Armstrong: 

 and Dr. J. V. Eyre. — The thermal conductivity of air and 

 other gases : G. W. Todd. The paper gives an account 

 of a determination of the thermal conductivities of air 

 and other gases at atmospheric pressure. The conductivity 

 was obtained from observations of the steady flow of heat 

 between two horizontal circular metal plates maintained 

 at different temperatures, the upper one at the temperature 

 of steam and the lower one at room temperature. The 

 upper plate was fixed, and the lower one could be mover! 

 uo and down so as to vary the distance between them. 

 If the temperatures of the plates are kept constant, the 

 quantity of heat passing per second from the upper plate 

 to the lower, when the distance between them is x, is 

 given by Q = K/.v4-R-(-E.y, where the constant K is pro- 

 portional to the thermal conductivity, R is the heat: 

 radiated, and E.v is the effect due to the edge. The latter 

 Is negligible when x is small compared with the radius of 

 the plates, so that the relation between O and x is given 

 by a rectangular hvoerbola. Hence the relation between 

 O and I 'x is a straight line the slope of which gives K, 

 from which the conductivity is determined. The inter- 

 section of this line with the .axis of O gives the value R 

 of the radiation. The value of the conductivity so obtained 

 was independent of the nature of the surfaces of the plates, 

 and also independent of the dimensions of the plates, the 

 latter nrovinir that convection currents were absent o- 

 negligible. The conductivities of some gases other than 



