550 



NATURE 



{Oct. 6, 1887 



flame is concealed by this chimney, excepting where a narrow- 

 slit, 10 X 3 mm., on each side shows the tip of the flame and 

 enables its height to be regulated. Through the interval between 

 the two chimneys the flame shines, and the light which it gives 

 is the same whenever the tip of the flame is visible through the 

 slit, whether towards the lower or the upper end. The two 

 chimneys are attached together by two curved metal bands 

 sufficiently removed from the flame on either side not to aifect 

 it. The attachment of. the bands to the lower chimney are 

 adjustable, so that the opening through which the central parts 

 of the flame are seen may be made larger or smaller. By means 

 of small cylindrical blocks, whose thickness is accurately gauged, 

 the width of the opening may be set either to that at which the 

 light emitted is one candle, or, if a greater or smaller light is 

 desired, a candle and a half or half a candle. The liquid with 

 which the lamp is fed is pentane, obtained in a manner already 

 described from American petroleum. 



Mr. W. N. Shaw read a paper by Mr. J. T. Bottomley on 

 Expansion by Heat of Wires under Pulling Stress. The wires 

 were two fine copper wires. One of them carried about half its 

 breaking weight and the other about a tenth of its breaking 

 weight. The wires were suspended in a tube, a scale being 

 attached to one, and a pointer moving over the scale to the 

 other. Thermometers were inserted into the tube at various 

 points, and the wires were heated by passing steam into the 

 tube. It was found that the more heavily weighted wire 

 extended much more than the lightly-weighted one. An amount 

 of permanent elongation remained, but more in the heavily- 

 strained wire. Each time the heating was done there was more 

 and more permanent elongation, and ultimately one of the wires 

 was broken under less than its breaking load in the normal 

 state. Further experiments were made with wires which had 

 been hardened, and the final result is that the coefficient of 

 expansibility for heat of copper wire strained by a certain 

 weight is greater than that of similar wire less heavily 

 weighted. 



Experiments on Electrolysis and Electrolytic Polarization, by 

 W. AV. Haldane Gee, Henry Holden, and Charles H. Lees. — 

 This is a preliminary notice of experiments that are in progress in 

 the Owens College Physical Laboratory. The experiments fall 

 under four heads : {A) electrolysis under pressure ; {B) time- 

 rate of fall of polarization in closed circuit ; (C) irreciprocal 

 conduction ; (Z>) the production of an oily fluid in electrolysis 

 with palladium electrodes. 



A. Numerous experiments have been made in order to deter- 

 mine the variation of the resistance of polarization of a sealed 

 voltameter in which dilute sulphuric acid was electrolyzed be- 

 tween platinum wire electrodes, it being thus subjected to the 

 pressure (up to 200 atmospheres) of the evolved gases. It was 

 found that the resistance markedly decreased, and the polariza- 

 tion decreased slightly. These changes may, however, it is 

 thought, be due to change of temperature, the influence of which 

 would appear from later experiments not to have been fully 

 eliminated. In two cases no change whatever was perceived : 

 (i) when two platinum plates were used as electrodes, and 



(2) when two voltameters were connected together forming a 

 sealed vessel, one voltameter being used to increase the pressure, 

 while observations were made on the other. As it has not been 

 possible to obtain glass tubes sufficiently strong for the high 

 pressures desired, an apparatus of gun-metal has been constructed. 

 This apparatus, which is fitted with a Bourdon's gauge recording 

 to six tons on the square inch, may also be arranged for pressure 

 experiments in general by attaching to it, by means of a strong 

 metal tube, a suitable receiver. In two of the experiments, when 

 the pressure had reached between 200 and 300 atmospheres, the 

 evolved oxygen and hydrogen gases combined with explosion, 

 although precautions had been used to prevent the gases from 

 coming into contact with the platinum, except in the liquid. 



B . The object of this research was to try to learn the parts 

 played by the various portions of the evolved gases : ( i ) that 

 occluded by the electrodes ; (2) that deposited on them ; and 



(3) that contained in the liquid in influencing polarization. The 

 method employed was to vary the conditions under control, 

 e.g. time of changing, density of current, &c., and to observe 

 the time-rate of the fall of the polarization thus produced in 

 closed circuit. It was found to be very difficult to apply this 

 method, because though the conditions under control were kept 

 2.S constant as possible, yet the time-rates of fall in two successive 

 observations were often different. This was thought to be 



due to the insufficient cleaning of the electrodes between each 

 experiment, and various methods were tried to remedy it, with 

 the general result that the more perfect the cleaning became the 

 more regular did the curves giving the time-rate of fall of the 

 polarization become, but still the inconsistencies were not wholly 

 removed. Heating the electrodes by the electrical current seemed 

 preferable to the other methods of heating. 



C. Whilst electrolyzing strong sulphuric acid between platinum 

 electrodes, it was noticed that when the current density at the 

 anode had exceeded a certain value decomposition apparently 

 ceased. The value of the anode current-density necessary to 

 produce this phenomenon is increased by diminishing the concen- 

 tration or increasing the temperature of the acid (thus dimin- 

 ishing the viscosity), and is diminished by cleaning the electrodes. 

 It was found that this great diminution of the current was not 

 caused by the formation of an opposing E. M. F., but by a 

 sudden increase of from 500 to 50,000 ohms in the resistance of 

 the circuit. That the insulating condition occurs at the anode 

 is shown by successively replacing the kathode and the anode by 

 clean plates ; in the first case the stoppage of the current per- 

 sists, in the second case the current is readily conducted. The 

 cause seems to be a sheath of oxygen bubbles which firmly 

 adhere to the anode when the insulating condition is formed. 

 The film is removed by momentarily breaking the circuit, or 

 short-circuiting the voltameter, or reversing the current, or by 

 replacing the anode by a clean plate. 



D. During the electrolysis of various liquids between palladium 

 electrodes it has been observed that a dense-looking liquid 

 streams from one of the electrodes (the anode in dilute sulphuric 

 acid, the kathode in caustic soda) after a reversal of the current. 

 The liquid seems to be a compound of oxygen and hydrogen, 

 presumably hydroxyl. 



On the Vortex- Theory of the Luminiferous Ether, by Prof. 

 Sir W. Thomson, F.R.S. — "In endeavouring to investigate 

 turbulent motion of water for my communication on that 

 subject to this Section, I have found a solution (many times 

 tried for within the last twenty years) of the problem — to con- 

 struct, by giving vortex-motion to an incompressible viscid fluid, 

 a medium which shall transmit waves of laminar motion as the 

 luminiferous ether transmits waves of light. Let xav, xzav, 

 xyzav denote space-averages, linear, surface, and solid, through 

 infinitely great spaces." After defining and illustrating this 

 method of averages by examples, and remarking in passing that 

 a general property of it is that 



xav —i^ = o, 



dx 



where Q is any quantity which is finite for infinitely great values 

 of X, he proceeded thus : — 



Suppose now the motion to be homogeneously distributed 

 through all space. This implies that the centres of inertia of all 

 great volumes of the fluid have equal parallel motions, if any 

 motions at all. Conveniently, therefore, we take our reference- 

 lines, OX, OY, OZ, as fixed relatively to the centres of inertia 

 of three (and therefore of all) centres of inertia of large volumes ; 

 in other words, we assume no translatory motion of the fluid as 

 a whole. This makes zero of every large average of u, and of 

 V, and oixv ; u, v, and w being the velocity-components; and 

 we may write as the general expression for nullity of translational 

 movements in large volumes — 



o = ave. u = ave. v = ave. w, 



where ave. denotes the average through any great length of 

 straight or curved line, or area of plane or curved surface, or 

 through any great volume of space. In terms of this general- 

 ized notation of averages, homogeneousness implies — 



ave. u- — U"^, ave. v^ — V^, ave. w^ = W^, 

 ave. vw = A^, ave. wu = B-, ave. uv — C'^, 



where U, V, W, A, B, C are six velocities independent of the 

 positions of the spaces in which the averages are taken. These 

 equations are, however, infinitely short of implying, though 

 implied by, homogeneousness. 



Suppose now the distribution of motion to be isotropic. This 

 implies, but is infinitely more than is implied by, the following 

 equations in terms of the above notation, with further notation, 

 R, to denote what we shall call the average velocity of the 

 turbulent motion — 



U2 = V2 = W- iR2, 



o = A = B = C. 



