378 RECORD OF SCIENCE FOR 1886. 



greatest errors arise from the liquid failing to be saturated with zinc 

 sulphate, in which case the E. M. F. is too high. The opposite error of 

 supersaturatiou is also sometimes met with. If these errors are avoided, 

 as may easily be done; if the mercury be pure, (profusely distilled m 

 vacuo); and if the paste be neutralized either originally with zinc car- 

 bonate or by allowing a few weeks to elapse (during which the solution is 

 supposed to neutralize itself), the electromotive force appears to be trust- 

 worthy to one-thousandth part. This conclusion is based on the exam- 

 ination of a large number of cells prepared by the author and by other 

 physicists. (:Nrature, February, 1886, xxxiii, 357.) 

 Lippmann has devised an absolute spherical electrometer, which con- 

 sists essentially of an insulated metallic sphere raised to the potential the 

 value of which we desire to know. This sphere is so constructed as to 

 divide into two hemispheres, which are movable with respect to each 

 other, and which, when the system is electrified, repel with a perfectly 

 definite force. It can easily be shown that this force is one-eighth of the 

 square of the potential of the sphere. To measure this force, one of the 

 hemispheres is fixed, the other is suspended by a tri-filar system — three 

 vertical wires of equal lengths. When repulsion is produced the movable 

 hemisphere can be displaced only i)arallel to itself: the three wires then 

 make a small angle with their original vertical position, which augle is 

 measured by the mirror method. The product of the weight of the 

 movable hemisphere by the tangent of the angle of rotation is the 

 force. In the improved form of this instrument the two hemispheres 

 are contained within a concentric spherical envelope, which is connected 

 with the earth. By this means not only is the sensitiveness increased, 

 but the system is protected from air currents as well as from extrane- 

 ous electrical influences. (C. E., March, 1886, cii, 6G6 ; Phil. Mag, July, 

 1886, V, XXII, 79; J. Phys., July, 1886, II, v, 323.) 



Bichat and Bloudlot have described two forms of absolute electrom- 

 eter, the one adapted to give continuous indications the other for high 

 potentials. In the former instrument, the attraction between two con- 

 centric cylinders is measured. One of these is fixed, the other is at- 

 tached to the beam of a balance; the other end of the beam carries a 

 disk moving in a cylinder to damp the oscillations. In the other elec- 

 trometer the movable cylinder is suijported at the middle of its length 

 upon a knife-edge attached to the beam ; it is balanced by adjustable 

 counterpoises. The upper portion of this cylinder is surrounded by a 

 second and hollow cylinder, which when electrified attracts the inner 

 cylinder upward. The lower end of the cylinder carries a scale-pan and 

 a damping cylinder of paper moving in a glass cylinder of slightly larger 

 diameter (C. R., March, cii, 753 ; July, 1886, cm, 245 ; J. Phys., II, V, 

 325, July ; 457, October.) 



Kolacek has suggessed a method of using the gold-leaf electroscope 

 so as to get a dednite relation between the augle of divergence of the 

 leaves and the difference of i)oteutial between the leaves and the enve- 



