November 14, 1901] 



NA TURE 



47 



the Hehaviour of Dielectrics," by Dr. J. Buchanan, was read by 

 the secretary. The action of this model depends on the 

 viscosity of a liquid, and the diagrams derived from it show by 

 their form that the motion of the pencil which traced them 

 approximated closely to what may be e.vpressed by the term 

 " motion of a viscous fluid by diffusion. " In other words, the 

 displacement curves obtained from the model and their derived 

 velocity curves are of the same form as the graphs of certain 



dv d-'ci 

 solutions of Fourier s well-known equation tt = K i-:n. Lord 



Kelvin has shown that the potential and the current at any 

 point in the wire of a calile can be expressed by appropriate 

 solutions of this equation, and in the same manner by the use 

 of solutions of this equation the diffusion of electricity into or 

 out of the dielectric of a condenser can be treated. It appears, 

 therefore, that the motion of the model and the diffusion of 

 electricity in a dielectric are subject to one and the same 

 mathematical law. The author suggests that the inventors should 

 obtain hysteresis diagrams by cyclical loading of the springs. 

 Prof. J. A. Fleming said he was glad that Dr. Buchanan had 

 drawn attention again to the model because there were points 

 about it which might be amplified with advantage. After giving 

 a short description of the apparatus he said that Dr. Buchanan 

 had shown that mathematically the theory of the model was 

 the same as that of diffusion in a cable, and he suggested that 

 there might be something more than mathematical analogy. 

 Prof. P'leming referred to the discussion on the original paper 

 in which Prof. Ayrton asked in what respect the model served 

 its purpose better than a twisted wire. A twisted wire cannot 

 represent the properties of a dielectric, because if twisted 

 beyond the elastic limits there is a permanent set. There is 

 no permanent set in the present model. He would like to 

 know if a dielectric has a true conductivity, and suggested that 

 experiments should be made by subjecting a dielectric to constant 

 electric pressure at constant temperature, for years if neces- 

 sary, and observing whether the curve of current becomes 

 asymptotic to the zero line or to a line parallel to it. The 

 model could be made to represent a conduction as well as a dis- 

 • placement current by so arranging the bottom piston that it 

 could descend but not return. The fact that the movements 

 of the model were similar to the diffusion of current in a cable 

 suggested that the process of conduction in a metal was similar 

 to that of displacement in a dielectric — Mr. [. Macfarlane Gray 

 read a note on the numerical value of the "characteristic" 

 of water. The author referred to a paper on thermodynamics 

 which he wrote twenty years ago and in which he supported 

 the theory of a granular ether under enormous pressure. This 

 theory easily explains the properties of bodies. There is a 

 numerical characteristic for every substance in the state of vapour. 

 This characteristic can be deduced from an analytical expression 

 involving certain physical data which must be experimentally 

 determined. His original number for water was 25-30693, 

 but later experiments by Lord Rayleigh on the weight of 

 hydrogen have altered this number to 25-33776. The author's 

 original value for the absolute specific heat of water was 

 124960 " mms. lift at Paris," but recent experiments of 

 Callendar give 126230. According to the author's theory, water 

 commences to freeze at 95' F. and the variation of the specific 

 heat of water at low temperatures is due to the latent heat of 

 ice. The formation of ice particles also explains the peculiar 

 changes in volume of water as it cools to the freezing point. 

 The chairman asked if this theory could explain the fact that 

 water can remain liquid below 32' F. Mr. Macfarlane Gray 

 said it could. 



Paris. 

 Academy of Sciences, November 4.— M. Bouquet de la 

 Grye in the chair. — On Analysis silus, by M. H. Poincare.— 

 On some chemical effects produced by the radium radiation, by 

 M. Henri Becquerel. It is pointed out that the radium radia- 

 tions consist ot a part capable of deviation in the magnetic 

 field, identical with the kathode rjtys, and a part non-deviable, 

 a fraction of which is absorbable and the remainder extremely 

 penetrating. Some kind of spectrum analysis is, therefore, 

 necessary before studying the chemical action of these rays. 

 Fresh observations are brought forward showing the action of 

 the rays upon glass, the transformation of yellow into red phos- 

 phorus, the reduction of mercury perchloride in the presence 

 of oxalic acid and the effect upon seeds. In the latter case it 

 was found that prolonged exposure to the radium radiations had 

 NO. 1672, VOL. 65] 



the effect of destroying the power of germinating in the seed. — 

 The electrolysis of ammonium chloride in solution in liquefied 

 ammonia, by M. Henri Moissan. Liquid ammonia at -So" C. 

 is readily electrolysed with a potential difference of 115 volts, 

 and it is remarkable that no nitrogen is produced. At the 

 positive pole chlorine is evolved, and at the negative pole 

 hydrogen, the purity of the latter being proved by analysis. 

 Dry iodine is not attacked or dissolved by liquid ammonia at 

 — 70°, or at temperatures below this, hut at higher temperatures 

 the iodine goes into solution. — The decomposition of calcium- 

 ammonium and of lithium-ammonium by ammonium chloride, 

 by M. Henri Moissan. Both calcium-ammonium and lithium- 

 ammonium react upon ammonium chloride in solution in liquid 

 ammonia at a temperature of -80° C. Under these conditions 

 the group ammonium could not be isolated, ammonia and 

 hydrogen being set free. — On a new method of detecting very 

 small electric charges, by M. R. Blondlot. Attempts were 

 made, without success, to determine some very small electric 

 charges by means of the usual electroscopes and electro- 

 meters. A new instrument was, therefore, constructed, de- 

 tails of which are given, possessing the required sensi- 

 bility. — The sugars in the blood and their glycolysis, by 

 MM. Lepine and Boulud. It is shown that the difference 

 between diabetic blood and normal blood consists not only in 

 the fact that the former preserves its reducing power better than 

 the latter, but also in the decisive fact that after keeping for an 

 hour in glass vessels at 39", the fermentable sugar of the blood 

 is not modified, whilst it is destroyed in normal blood. — Remarks 

 by M. Marey on two reports on chronophotography and of a 

 commission on physiology and hygiene. — Report by a committee 

 appointed to examine the papers left by the late M. Halphen. 

 The memoirs left in a state fit for publication are too few in 

 number to publish in volume form, but it is desirable that some 

 periodical would insert certain fragments. — Observations of the 

 1901 comet made at the Observatory of Santiago, Chili, and the 

 elements of the same comet, by INL Obrecht. — Sunspots and 

 planets, by M. Birkeland. The results given in a former paper 

 have been recalculated, taking into account the action of the 

 planet Saturn, but the conclusions previously arrived at are not 

 thereby altered. — On persistent conjugated network, by M. J. 

 Raffy. — On the adiabatic curve, by M. George Moreau. The 

 usual equation to the adiabatic curve, PV'' = const, is obtained 

 under the suppositions that the ratio of the specific heats, the 

 specific heat at constant volume and the coefficient of expansion 

 are const.ant. It has been shown, however, by MM. Mallard 

 and Le Chatelierthat the specific heat at constant volume is not 

 constant, but is a linear function of the temperature, and the 

 coefficient of expansion is also a function of the temperature. 

 On these assumptions a more general form of the equation to 

 the adiabatic curve is worked out. — On the chlorobromides of 

 thallium of the type Tl4Xg, by M. V. Thomas. The current 

 theories of the constitution of double salts would allow of the 

 prediction of two sesquichlorobromides of thallium. The mode 

 of preparation of two isomers of TljCl3Br3 is described in detail. 

 — Some reactions of trichloracetic acid, by M. A. Clermont. 

 The ethyl ester and amide of this acid are so readily prepared 

 that their formation may be used as tests for the acid. 

 — Researches on some isomerides of pinacone and its deri- 

 vatives, by M. Maurice Delacre. The reactions of pinacoline 

 agree in part with thp formula suggested by Butlerow 

 (CHjjj.C.CO.CHj, and in other respects corresponds to 



Friedel's formula (CHjjoC C— (CHj),. As a result of the ex- 

 periments here given the author inclines to the view that 

 pinacoline contains the two substances represented by the 

 above formulae in a state of equilibrium. — The constitution of 

 piceol, by MM. Ernest Charon and Demetrius Zamanos. The 

 glucoside piceine, extracted by AL Tanret from Pinus picea, 

 was shown by him to be hydrolysed by acids into glucose and 

 a .substance piceol. It is now shown that this latter substance 

 is paraoxyacetophenone, the properties of the natural and syn- 

 thetical piceols agreeing completely. — On the calculation of the 

 amounts of water added to and cream abstracted from milk, by 

 M. V. Genin. — On the formation of the perfume of vanilla, by 

 M. Henri Lecomte. The following hypothesis would best ap- 

 pear to explain the formation of vanillin in the fruits during their 

 preparation. The coniferine is converted into coniferyl alcohol 

 and glucose by means of a hydrolytic ferment, crude vanilla, in 

 fact, always containing glucose. This alcohol is then trans- 

 formed into vanillin by the action of an oxydase, the existence of 



