Mahch 19, 1909] 



SCIENCE 



477 



To measure the torque one of the concentric ad- 

 justments is then made until the tone disappears 

 and the angular change read on a scale graduated 

 preferably to horsepower per revolution. The 

 pitch of the tone compared with a calibrated tun- 

 ing fork gives the speed. The necessary calibra- 

 tion of the shaft is made before it is placed in 

 position or it can be made afterward. (A model 

 was shown.) 



Some Optical Effects of Changes in Ether Density : 

 Charles P. Brush, Cleveland, Ohio. (Read 

 by title.) 



The Lumeter, a Practical Measure of General 

 Luminosity : Henet E. Wetheeell, M.D., Phila- 

 delphia. (Read by title.) 



A Ballistic Dynamometer Method of Measuring 

 Hysteresis Loss in Iron: Martin E. Rice and 

 BuETON McCoLLTJM, University of Kansas. 

 (Read by title.) 



The sample to be tested, which should be lami- 

 nated, is wound with a primary coil and a test 

 coil. The latter is connected in series with the 

 fine wire movable coil of a dynamometer and the 

 former in a series with the coarse wire-fixed coil 

 of the dynamometer. When the primary current 

 is reversed, the dynamometer measures by its 

 ballistic deflection the total energy loss per half 

 cycle in the iron and in the test coil circuit. 

 Since the hysteresis loss is independent of the 

 period of the cycle while all the other losses 

 measured are inversely proportional to the period 

 of the cycle, it is easily possible by the insertion 

 of a choke coil in the primary circuit and the use 

 of a moderately high resistance test coil circuit 

 to keep these other losses below one per cent, of 

 the total loss measured. A comparatively rough 

 estimate of these other losses is then sufficient to 

 enable the true hysteresis loss to be determined 

 with an error of only a small fraction of one per 

 cent. 



The dynamometer can be calibrated by dis- 

 charging a condenser or a mutual inductance 

 through its movable coil while a constant current 

 is maintained in its fixed coil. 



This method eliminates the difficulties inher- 

 ent in the wattmeter methods due to uncertain- 

 ties in frequency and wave shape while it avoids 

 the tedious process of taking a long series of 

 readings, plotting a curve and measuring its area, 

 as in the ballistic galvanometer methods. Tests 

 can be made much more rapidly than by the watt- 

 meter methods, while the results obtained are 

 fully as accurate as by the ballistic galvanometer 

 methods. 



On the Diurnal Variations in the Intensity of the 

 Penetrating Radiation Present at the Surface of 

 the Earth: G. A. Cline, Toronto University. 

 On the Character of the Radiation from Potas- 

 sium: J. C. McLennan, Toronto University. 

 The Action of Electrolytes on Copper Colloidal 

 Solution: E. F. Burton, Toronto University. 

 The experiments detailed in the present com- 

 munication are a continuation of those performed 

 by the writer on the action of small traces of 

 electrolytes on silver and gold colloidal solutions 

 prepared by Bredig's method. With these solu- 

 tions the particles in the pure solution are nega- 

 tively charged and it was found that, if an elec- 

 trolyte was added, the positively charged ion was 

 the potent one in reducing the velocity with which 

 the particle moved in a given electric field; i. e., 

 the ion charged oppositely to the colloidal par- 

 ticle produces the discharge of the particle and 

 consequently coagulation of the colloid. 



Copper colloidal solutions were chosen on 

 which to work because they have positively 

 charged particles in the pure solution. The elec- 

 trolytes used were solutions of potassium chlor- 

 ide, potassixmi sulphate, aluminium sulphate, 

 potassium phosphate, potassium ferricyanide. 

 With this series it was possible to compare the 

 eflTect of the monovalent and the trivalent ions of 

 both acids and bases. 



Every electroljrte added produced a decrease of 

 the velocity with which the copper particles 

 moved to the cathode. It is the ion bearing the 

 negative charge which is active in reducing the 

 velocity. This power of the negative ion depends 

 on the valency in a way analogous to the rela- 

 tions foxmd by Picton and Linder, and Hardy for 

 the coagulative power of ions. Evidence is also 

 produced to show that the discharging power of 

 two negative ions of the same valency is the 

 same. Current observations on the coagulation 

 of the colloids in each case showed that the par- 

 ticles coagulate more and more freely as the 

 charge gets smaller and smaller. 

 Arc and Spark Phenomena in the Secondary of a 

 High Potential Transformer : E. S. Johonnott, 

 Rose Polytechnic Institute. (Read by title.) 



The Upper Inversion in the Atmosphere: W. J. 



Humphreys, Mt. Weather Observatory, Md. 



We have been accustomed to think of the at- 

 mosphere as growing steadily colder to nearly or 

 quite absolute zero with increase in elevation, but 

 hundreds of records obtained in many parts of 

 the world by the aid of free balloons show that 

 this assumption is very wide of the truth. 



