[ 462 ] 



LXI. A Modification of Dumas's Method of Determining 

 Vapour -Densities, ^y J. Habermann*. 



AS the significance of the knowledge of the vapour-den- 

 sities of substances in relation to the doctrine of modem 

 chemistry is gradually being more fully recognized , chemists 

 are striving to simplify the methods of Gay-Lussac and Dumas 

 (the two chief processes at present in use for obtaining such 

 knowledge), and at the same time to render them more accurate 

 and of wider applicability. 



In the year 1862, H. SchifF, in describing a modification of 

 Gay-Lussac's method, says, " In spite of all the improvements 

 which have hitherto been proposed, chemists seem to prefer 

 to adopt the method of Dumas in making determinations at 

 low temperatures." At the present time, however, the opposite 

 appears to be the case. 



Since the time of Hofmann's improvement of Gay-Lussac's 

 method, the efforts of chemists towards perfecting the modes 

 of determining vapour-densities have principally been di- 

 rected to removing the disadvantages connected with the 

 Hofmann- Gay-Lussac method, and to making it applicable to 

 bodies which boil at a high temperature. As a result of these 

 investigations, we have existing at the present time no less than 

 three memoirs relating to the subject. 



The last publications treating of improvements of Dumas's 

 method were made some years ago by Deville and Troost, and 

 Bunsen. The object of the two first-named investigators was 

 to extend the limit of temperature to which the method is 

 applicable beyond the boiling-point of zinc. Bunsen so varied 

 the process that by his modification it is possible to obtain the 

 vapour-densities of bodies by a determination depending 

 merely on two weight-differences ; and this can be done with a 

 degree of accuracy hitherto hardly known. The care and 

 attention needed in carrying out this method, however, are 

 increased in a ratio corresponding to this accuracy ; and in 

 consequence the modified process up to the present time has 

 not been very generally adopted, more especially as it fails to 

 remove certain important disadvantages of the original method. 

 Later suggestions for improvement in this direction are entirely 

 wanting ; and in consequence, whenever practicable, the method 

 of Dumas is dispensed with, since, to quote Victor Meyer, ^' it 

 necessitates a loss in material of about three grams, and is there- 

 fore not easily applicable to the greater number of newly disco- 

 vered bodies." Moreover, since it requires a temperature greatly 

 exceeding the boiling-point of the substance, the determination 

 * Freely translated from Liebig's Annalen, June 9, 1877, by John I. Watts. 



