158 RESEARCHES ON THE VOLATILE 



elementary atoms which it contains, without deducting from this sum a constant 

 number, as by Schroder's method. According to Lo wig's theory, one atom of carbon 

 (C) raises the boiling-point 38°.4, and one atom of hydrogen (H) lowers it 29 3 .2 ; 

 these numbers being for carbon nearly two and one half times, and for hydrogen 

 nearly three times as great, as those of Schroder. Gerhardt,* in a special paper 

 " On the Boiling-point of the Hydrocarbons," observes that " The boiling-point of the 

 hydrocarbons appears to obey a very simple law, according to which it is raised or 

 depressed a certain number of degrees, corresponding to the number of equivalents 

 of carbon or hydrogen contained in its equivalent."! From a comparison of the 

 boiling-points and formulae of several well-known hydrocarbons, the determinations of 

 which were repeated with special care for this purpose, Gerhardt finds that the addi- 

 tion of C 2 to the molecule of an hydrocarbon raises its boiling-point 3 5°. 5, and that 

 the addition of H 2 lowers it 15°. The boiling-point of a body is calculated from these 

 numbers by comparing its formula with oil of turpentine, C 20 H 16 , as a standard, the 

 boiling-point of which is taken at 160° C. Example : cumole (from cuminic acid) has 

 the formula C 18 H 12 ; hence it contains C 2 less than oil of turpentine ; therefore 35°.5 

 must be deducted from 160° (the boiling-point of oil of turpentine), which leaves 

 124°.5 ; but as the cumole contains 2 H 2 less than oil of turpentine, 15° X 2 = 30° is 

 to be added to the above remainder ; thus 124°.5 -f- 30° = 154°.5, the calculated 

 boiling-point of cumole. Gerhard t's direct determination was 153°, which very nearly 

 coincides with his theory. 



It would be foreign from my purpose on the present occasion to consider these 

 different hypotheses, or even the empirical law of Kopp, beyond their special relation 

 to the boiling-points of the hydrocarbons, and such other series, derivatives from the 

 hydrocarbons, as have been made the subjects of my own experiments. Anything 

 more than this would be merely speculative. The want of more accurate determina- 

 tions of boiling-points as essential to safe and reliable deductions and generalizations 

 on this question, has frequently been observed. The need of this will be made 

 strikingly apparent by comparison of my results with those of previous observers. 



Indeed, if my determinations may be taken 



which, considering the 



of the materials operated upon, might not be quite fair 



of 



the boiling-points which have hitherto been published are probably so numerous, and 



* Annales de Chiraie et de Physique, 1845, 3 e Serie, XIV. 107. 



t " II parait que le point d'^buUition des hydrogene carbon^ est soumis a une loi fort simple, d'apres laquelle 

 il s'eleverait ou s'abaisserait d'un certain nombre de degr& suivant le nombre des equivalents de carbone ou 

 d'hydrogene renfermes dans leur equivalent," 



