1891.] On certain Ternary Alloys. 181 



verified in practice, or whether interfering causes prevent anything 

 more than demonstrations of approximate correctness being obtained ; 

 the more so that some of the results previously obtained by two of us 

 do not appear to be in harmony with Sir G. G. Stokes's proposition. 

 In Part I (' Boy, Soc. Proc.,' vol. 45, p. 461) three series of experi- 

 ments were described, made with lead, tin, and zinc, where the ratio 

 of lead to zinc was 2 to 1, 1 to 1, and 1 to 2, in the three series 

 spectively ; and the figures obtained led us to the conclusion that, 

 rhilst an indefinite number of different mixtures may be prepared, 

 one of which will give the same heavier alloy, the lighter alloy 

 simultaneously formed will be different in each case; and conversely: " 

 a deduction obviously incompatible with Sir G. Gr. Stokes's proposition. 

 On the other hand, it is argued by Sir Gr. G. Stokes that these experi- 

 ments do not necessarily prove anything more than the extreme 

 difficulty experienced whilst making experiments with fused metals 

 in obtaining such an intimate intermixture as to bring about the 

 condition of perfect equilibrium between the two alloys formed in 

 any given instance ; and that, in point of fact, the differences 

 observed in the compositions of the various lighter alloys associated 

 with a given heavier one, or vice versa, are not greater than might 

 reasonably be expected were equilibrium not perfectly attained in 

 some or all of the observations. Further, the fact that the differences 

 are always in the same general direction tends to indicate that some 

 constant interfering cause is at work ; thus, when curves were 

 plotted (Part I, fig. 5, p. 476) with the tin percentages in the heavier 

 alloys as abscissae, and the excesses of tin percentage in the lighter 

 alloys over those in the heavier ones as ordinates, the curve deduced 

 from the series of experiments where the ratio of lead to zinc in the 

 original mixture of metals was 2 to 1 underlay that similarly obtained 

 in the second series, where the ratio was 1 to 1, which again underlay 

 that deduced from the third series, where the ratio was 1 to 2 ; 

 whereas all three curves should have coincided were Sir G. G. Stokes's 

 proposition correct, and all interfering causes completely eliminated. 



An analogous result is obtained when the analytical figures are 

 plotted on Sir G. G. Stokes's triangular system. Fig. 3 represents 

 the plottings thus obtained of the two series where the ratio of zinc- 

 to lead was 2 to 1 and 1 to 2 respectively (Part I, 'Boy. Soc. Proc.,' 

 vol. 45, Series IV, p. 474, and Series VI, p. 475) the temperature 

 throughout being near to 650. The ties in the first case are indi- 

 cated by dotted lines, and in the second by continuous ones. 

 Obviously the critical curves deduced from the two sets of observa- 

 tions respectively do not differ very markedly ; but the angles of 

 slope of the ties are not identical, so that a given heavier alloy is not 

 conjoined with the same lighter one (nor vice versa) in the two cases ; 

 whilst the direction of the variation is the same throughout. 



