Intelligence and Miscellaneous Articles. 135 



2 per cent. After trying many plans one was finally adopted that 

 gives individual results that agree to about g per cent. ; and the 

 average departure of single observations from the mean of several 

 seldom exceeds \ per cent. 



The waves were generated by a fork whose frequency was often 

 determined, and was always near 62-87 double vibrations per 

 second. The water and solutions were contained in a porcelain 

 tray 1 by 12 by 14 inches. The wave-length was measured by 

 means of a telescope mounted on a dividing-engine, whose screw 

 had a pitch of 1-0328 milliui. The waves were invisible under 

 ordinary conditions, but were observed by Foucault's method for 

 rendering visible small vibrations in plane or spherical surfaces. 



The water used was especially distilled by Mr. W. T. Mather from 

 chromic acid and alkaline potassium permanganate, and was con- 

 densed in a block-tin condenser ; it was the kind used by him for 

 his electrolytic work. The salts were obtained from Eimer and 

 Amend, and were said to be chemically pure. 



With this apparatus 1 have determined the surface-tension of 

 water and of solutions of sodium chloride, potassium chloride, 

 sodium carbonate, potassium carbonate, and zinc sulphate, of con- 

 centrations varying from O05 normal to normal. 



The value found for water is T = 75-98 dynes per centimetre at 

 0° G. ; while Seutis (Journ. de Phijs. (3) vi. p. 183, 1897), working 

 by an entirely different method, found T = 76-09 at 0° C, which 

 differs from the other by only 0-14 per cent. These values agree 

 very well with the values given by Lord Rayleigh, Hall, Volkmann, 

 and others, but are much lower than Quincke's value. 



It was found that the surface-tensions of dilute aqueous solutions 

 are linear functions of the concentration : so that we may write 

 T i . = T w -|-^C, where T s = surface-tension of the solution, T M .= sur- 

 face-tension of water at the same temperature, kis a constant, C is 

 the concentration in gramme molecules per litre. Below is a table 

 showing the values of k as determined by different observers : — 



Dorsey. Volkmann. Quincke. Bother. 



XaCl 1-53 1-59 1-57 1-38 



KC1 2-23 1-41 1-57 1-47 



ijS T a 2 C0 3 2-00 0-987 1-57 



|K 2 C0 3 1-77 1-78 1-57 



ZnS0 4 1-86 



Volkmann found that the curve for ]\a 2 C0 3 at great dilution 

 becomes steeper than the one for K 2 C0 3 , which agrees with my 

 results. Quincke's value 1-57 does not agree with his results 

 except for KC1 and NaCl. I cannot account for the very high 

 value I found for KC1 ; but it must be borne in mind that the 

 values given above are not fairly comparable, since my values are 

 for solutions generally less concentrated than | normal, while the 

 other are found for solutions of greater concentration. — Johns 

 Hooking University Circulars, June 1897. 



