TRANSACTIONS OF SECTION A. 80.3 



A series of observations thus plotted wouM enable one to determine whether 

 <he actual behaviour of the mixture at hi^'her pressures tended to approach to agree- 

 ment with Dalton's law or some modificatiou of it such as that suggested. 



Regnault's observations include a series of determinations of the pressure and 

 related volumes of mixtures of air and ether vapour for the temperature 77° C. 

 They also show the dry-air pressures for the series of observed volumes. A 

 diagram showing two sets of his observations was exhibited. The pressures varied 

 between 600 and 1,400 millimetres. It was pointed out that for the observations 

 at greater rarefactions the Boyle's law line was very strictly adhered to until the 

 pressure reached 80 per cent, of saturation, then slight discrepancy in the direction 

 of loss of pressure was indicated, increasing until it amounted to upwards of 10 

 millimetres for tlie theoretical saturation point ; beyond that the discrepancy showed 

 a slightly diminishing value. A straight line drawn to show a theoretical defect 

 from Dalton's law amounting to 'Kp, where X = '007, agreed much more satisfactorily 

 with the plotted observations. 



The second series of observations made with a greater amount of ether showed 

 the discrepancy from Dalton's law gradually vanishing with increased pressure. 



It was pointed out that in drawing his conclusions Kegnault had not taken 

 account of the supersaturation of air which would result from compressing air even 

 at a constant temperature in the absence of nuclei for condensation and the time 

 that would be required for the gradual deposition of the excess of moisture upon 

 the liquid surfaces exposed to it. 



Moreover, it seemed most unlikely that if the presence of air caused a dis- 

 crepancy of 10 mm. for a total pressure of 1,000 mm. the increase of the pressure 

 to 1,400 mm. should do away with the discrepancy previously caused; in other 

 words, the line connecting the observations and intersecting the theoretical line at 

 the highest pressure recorded is antecedently improbable and unreal. 



The conclusion drawn was that Regnault's explanatii m is probably an unreal 

 one, and that an actual divergence from Dalton's law is indicated. The amount of 

 the divergence, however, cannot be finally deduced from Regnault's observations, 

 and must wait for a repetition of those or similar experimi-nts, in which the errors 

 which may be due to supersaturation are guarded against. 



6. On the Determination of the State of lonisation in Dilute Aqueous Solu- 

 tions containing two Electrolytes : No. 2. By Professor J. G. 

 MacGeegor, Dalhousie College, Halifax, N.S. 



In the Report of last year methods were described ol determining the lonisa- 

 tion coefficients in solutions containing two electrolytes, with or without a 

 common ion, and with no mutual chemical action other than double decomposition ; 

 and a statement was given of tests of accuracy which had been applied to the 

 coefficients thus obtained, by the employment of them for the calculation of the 

 conductivity of the solutions. Since that date other tests have been applied by 

 students in my laboratory, as follows: — E. H. Archibald has shown that the 

 conductivity of solutions containing K^SO^ and Na^SO^ can be calculated within 

 the limits of experimental error up to a total concentration of about 1 gr.-eq. per 1. 

 T. 0. McKay has obtained the same result for solutions containing BaOl,, and 

 NaCl. Archibald has completed his observations (referred to last year) on the 

 conductivity of solutions containing NaOl and K^SO., (and therefore also KCI and 

 Na-^SO^) ; and they show that the conductivity is similarly calculable up to a total 

 concentration of about O'o. Archibald has also shown that if the variation, with 

 concentration, of the surface tension and specific gravity of simple solutions of 

 K.jSO^ and Na.iSO^ be known, the s. t. and sp.gr. of mi.xtures up to concentrations 

 of about 1 and 0-5 respectively can be calculated, the method employed being that 

 of my paper in the ' Phil. Mag. ,' vol. 44 ; and that with similar data as to the 

 sp. gr. of NaCl, KCI, Na^SO^, and K,jS04 solutions, the sp. gr. of solutions con- 

 taining all these salts up to a concentration of about 0'8 is calculable. It should 

 be noted also that calculations made, by the aid of these coefficients, on the 



3f2 



