866 



SCIENCE. 



[N. S. Vol. XV. No. 387. 



this date being chosen instead of the usual 

 one on account of the reception tendered on 

 that day to the Society's only honorary mem- 

 ber. Lord Kelvin. The meeting was the most 

 largely attended in the history of the Society. 

 The program, which was so extended as to 

 make it hardly practicable to give an abstract 

 of the individual papers, was as follows : 



' Note on the Specific Heat of Mercury ' : H. T. 

 Baknes. 



' On tlie Theory of Concentration Cells ' : H. S. 

 Carhakt. 



' An Apparatus for the Quantitative Study of 

 Sound ' : A. G. Webstee. 



' The Magnetic Deviation of Rays from Radio- 

 Aetive Substances ' : E. Rutherford and A. G. 

 Grdse. 



'The Condensation of Nuclei': Carl Barus. 



' A New Gravity Electrical Time Key ' : Chas. 

 Forbes. 



' An Electrical Method for Calibrating Chrono- 

 graphs ' : H. C. Parker. 



'Absorption Curves for Condensers for Very 

 Short Time Intervals ' : H. C. Parker. 



' Residual Magnetism in Iron and Steel for Very 

 Short Intervals of Time ' : C. C. Trowbridge. 



'An Experiment Relating to the Application of 

 ijagrange's Equations to Electric Currents' : W. 



8. DAT. 



' The Physical Meaning of Mathematical Opera- 

 tions in Heat Conduction ' : A. S. Mackenzie. 



'A New Method of Integrating one of the Dif- 

 ferential Equations of Heat ' : R. S. Woodward. 

 'An Instrument for Drawing a Sine Curve ': A. 

 S. Mackenzie. 



' Three Lecture Experiments ' : W. S. Franklin. 

 Ernest Meeeitt, 



Secretary. 



PHILOSOPHICAL society OF WASHINGTON. 



The 552d regular meeting was held April 

 26, 1902. Mr. W. J. Spillman, of the Depart- 

 ment of Agriculture, described 'A Machine for 

 producing Normal Equations from Observa- 

 tional Equations.' 



The work to be done in this case consists in 

 multiplying each equation through by the co- 

 efficient of a given term of the equation, and 

 in collecting the new coefficients of homologous 

 terms into sums which are the coefficients in 

 the normal equation. The method here de- 

 scribed is applicable to eases where the number 



of terms in tlie normal equation is not greater 

 than four. The machine contains a number 

 of levers, pivoted at center, and graduated 

 each way from the center to 10. The length 

 of the graduations is arbitrary. Each lever is 

 provided with four indicator slides, two on 

 each side. The slides are set so as to point out 

 on the lever scale the coefficients of the terms 

 of an observational equation. The lever is 

 then said to be set for that equation. Erom 

 the slides cords pass upward to four systems 

 of pulleys, in such manner that moving the 

 levers vertically causes each cord to actuate an 

 index finger attached to the uppermost pulley 

 in each system. A convenient scale is arranged 

 near one end of the lever, and is graduated 

 into ten equal divisions, numbered from to 

 10. The lever being raised from the zero to 

 any point on this scale causes the index fingers 

 to move on four similarly graduated arcs, and 

 to point out on these scales the products of the 

 coefficients by the number to which the in- 

 dicator on the end of the lever points. 



Other levers are set for other observation 

 equations, as above. When any lever is raised 

 so as to point to the number by which its equa- 

 tion is to be multiplied, the new coefficients 

 of that equation are added to those already pro- 

 duced. When all the levers that are set are 

 properly adjusted, the index fingers point out 

 the coefficients of the normal equation. 



A similar machine had proved very useful 

 in problems requiring multiplication and sum- 

 mation of corresponding products. 



The next paper was by Mr. L. J. Briggs, also 

 of the Department of Agriculture, 'On the 

 Absorption of Gases and Dissolved Salts by 

 Quartz and Glass.' 



In one series of experiments very finely 

 powdered quartz was used and the weights of 

 water vapor or carbon dioxide absorbed thereby 

 were determined; these were found to be closely 

 proportioned to the pressure of the vapor or 

 gas. Similarly experiments were made with 

 very dilute solutions of chlorides, carbonates 

 and hydroxides of sodium, potassium and 

 aminonium. The quantity absorbed increases 

 much less rapidly than the concentration and 

 appears with these salts to be dependent on 

 the acid radical rather than upon the basic ele- 



