96 



Scientific Proceedings, Royal Dublin Society. 



associated compounds the deviations diminish with rise of molecular weight 

 and eventually become small. 



In Table I the values given were obtained by taking the absolute boiling- 

 point of methane to be 106-75, and calculating the boiling-points of the 

 succeeding paraffins from the values of A given by the formula. 



During the last few months a series of papers has been published in the 

 Journal of the American Chemical Society, on the boiling-points and critical 

 temperatures of some carefully prepared and purified paraffins. The data 

 are tabulated below : — 



Absolute Boiling-points. 

 Normal propane, 229-0°, Burrell and Eobinson, . . xxxvii., 2188 (1915) 

 Normal butane, 272-8°, Burrell and Robinson, . . . „ „ „ 



Normal nonane, 423 , 5°, Latham Clarke, and Adams, . „ 2536 „ 

 Isobutane, . 259-7°, Burrell and Eobinson, . . . „ 2482 „ 



xxxvii., 2683 (1915) 



Absolute Critical Temperatures. 

 Normal butane, 426 - 2°, Siebert and Burrell, . . 

 Isobutane, . 406-7°, Siebert and Burrell, . . „ „ „ 



The observed boiling-points of these paraffins differ somewhat from those 

 given by previous authors ; they are probably more accurate, and are adopted 

 in Table I instead of the values given in earlier tables. 



Taking the whole series from CH 4 to Ci 9 rl 40 , it is clear that the best results 

 are given by the formula 



A = 



144-86 



t, 0-0148^7" ' 



and that, of the other formulae, Ferguson's is available over a wider range 

 than either of the others. The relative availability of the formulae is shown 

 by the following comparison : — 



Number of differences smaller than 1° 

 Number of differences greater than 2°, 

 Greatest difference, .... 

 Greatest difference, omitting CH 4 , 



The boiling-points of only four members of the series of isoparaffins have 



been determined. These are given in Table II, together with the values of 



A derived from the observed boiling-points, those calculated from the formula 



144-86 



— , and the differences between the calculated and observed 



A = 



o-ousVr 



T 

 values of A 



