PHYSICS. 335 



sections of co-axial cylinders representing tlie isodynamic surfaces of 

 any compound pendulum. (Am. J. Sci., January, 1880, III, xxxi, 22.) 

 In determining density by means of the balance, it is desirable to 

 know to how many decimal places the division should be carried when 

 the error of weighing is the nth part of a grau). Leruiantoii" has in- 

 vestigated this question. Differentiating the ordinary expression for 



P dp dO 



the density D = — , we have dD = -— — D ^. Hence (1) an error of a 



certain fraction of a gram in the determination of the weight of a body 

 produces in the density obtained an error equal to the same fraction of 

 unity divided by the number of cubic centimeters occupied by the body ; 

 and (2) the influence of an equal error in the detennination of the 

 weight of water displaced is D times greater than that above given 

 and of contrary sign. (J. Soc. Phys. Chim. Russe, 1885, xvii, 56 ; J. 

 Phys., February 1886, II, v. 91.) 



Parize has suggested the following method of determining the density 

 of porous friable bodies, such as earth, peat, and the like. A jar of 

 250 cubic centimeters capacity is exactly filled with a smooth, regular 

 seed, such as linseed, for example, and is then weighed. The peat or 

 other porous material is placed in the jar and shaken down to its normal 

 condition, and the jar is again weighed. Knowing the weight of water 

 required to fill the jar, all the data necessary to calculate the density 

 are obtained. (J. Phys., May, 1886, II, v, 222.) 



Kirchhoff has calculated the change of form which an iron sphere 

 would undergo under the influence of a constant magnetic force. 

 Calling n the number of turns of wire in the magnetizing spiral and i 

 the current in amperes which traverses it, he flndsfor the elongation of 

 the radius of the sphere parallel to the axis of the spiral the value 

 n^ i^ . 2 . 32 . 10^^ . P. The contraction in a direction perpendicular to 

 this axis is between one-fifth and one sixth of this. (Wied. Ann., 1885, 

 XXIV, 52; XXV, 601; J. Phys., 1886; April, II, V, 175, 179.) 



Koch has published an account of experiments on the elasticity of 

 ice, made by him in Labrador in ]882-'83, and in Fribourg in 1884. 

 The co-efficient was determined from the flexure of bars of ice when 

 supported at the ends and weighted in the middle. The principal diffi- 

 culties encountered arose, first, from the direct evaporation which took 

 place, and which changed the dimensions of the bars ; second, from the 

 plasticity of the ice, which produced a permanent and progressive set ; 

 and third, from the existence of a consideraDle residual elasticity. Al- 

 lowing for these perturbing causes, the author obtained for the mean 

 value of the elasticity co-efficient, for ice cut parallel to the free surface 

 of solidification, the value 041.5, expressed in kilograms per square 

 millimeter; a value considerably higher than that obtained by Reusch 

 by the acoustic method, which was only 236.3. Koch, however, obtained 

 a still higher value by the acoustic method, namely, 884 kilograms 



