590 SCIENTIFIC RECORD FOR 1888. 



consequently the double refraction depends, certainly in soine 1 eases and 

 perhaps in all, on the modifications which the light undergoes in trav- 

 ersing the molecule. 2d, the action of heat on crystals may produce 

 three distinct classes of phenomena: (1) the axis of the ellipsoid of 

 elasticity may vary considerably in magnitude, these variations being 

 due, as in the case of boracite, to a change in the form of the molecule; 

 (2) the orientation of the molecules may vary suddenly, they being able 

 to turn about their centers of gravity so as to take various positions 

 compatible with their crystalline arrangement, which remains sensibly 

 constant, or is altered only by the very secondary phenomena of expan- 

 sion ; (3) the action of heat may change suddenly either the symmetry 

 of the molecule alone (boracite, potassium sulphate) or both this and 

 the reticular system. In case (1) one form passes into the other sud- 

 denly at the same temperature, the two not coexisting together. In 

 case (2) the change takes place at a fixed temperature only when the 

 temperature is rising ; when it falls a sort of crystalline surfusion takes 

 place, the form belonging to the higher temperature existing at a lower 

 but in an unstable condition. (J. Phys., May, 1883, II, n, 201.) 



Vieille has investigated the influence of cooling on the value of the 

 maximum pressures developed in a closed vessel by exploding gases. 



S 

 If the results are plotted in a curve, using v , or the ratio of the surface 



of cooling to the volume of the gaseous mass, as abscissas and the cor- 

 responding pressures as ordinates, this curve will be independent of 

 the nature of the walls of the vessel and of its capacity, and the point 

 of intersection of this curve with the axis of ordinates will give the 

 value of the pressure which the exploding gas would develop in an in- 

 closure impermeable to heat. The curves obtained are divisible into 

 two classes. The first, obtained with dissociable mixtures (CO and O, 

 H and O), are concave toward the axis of abscissas and tend to cut the 

 axis of ordinates at right angles. Extrapolation gives, then, a small 

 correction and the limiting pressure is accurately determinable. The 

 second, obtained by burning cyanogen and oxygen mixed with an inert 

 gas, are convex toward the axis of abscissas, and rise sensibly toward 

 the axis of ordinates. Hence the influence exerted by the unit of cool- 

 ing surface upon the diminution of pressure increases with rise of 

 temperature. The point at which the curve cuts the axis of ordinates 

 is less well defined in this case. {Comptes Rendus, January, 1883, xcvi, 

 11G.) 



Guthrie has described to the London Physical Society an experiment 

 which he had made on the theory of regelation. He modified Bot- 

 tomley's well-known experiment of cutting through a block of ice with 

 a metallic wire weighted at the ends, without separation of the frag- 

 ments, by using a cord of silk of the same size as the wire and equally 

 weighted. While the wire cut through the block the cord did not. 



