140 HISTORY OF CHEMISTRY. [LECTURE VIII. 



lich had studied the phenomena of isomorphism. It could then 

 be said that certain elements in a crystal might be replaced by 

 others, without alteration of the crystalline form. Such substi- 

 tutions possessed the peculiarity, however, that they were not 

 connected with any proportions by weight, and it may thus ap- 

 pear all the more remarkable that they should render important 

 assistance in the determination of atomic weights. The hypo- 

 thesis underlying the phenomena of isomorphism was that one 

 atom could only be replaced by one other ; that is to say, that 

 the numbers of the atoms in isomorphous compounds must be 

 identical. Since chemically similar substances had alone been 

 compared, an extension of the prevailing views, based on the 

 phenomena of isomorphism, would have been quite possible ; 

 but this class of phenomena had never led to any attack upon 

 the system. 



Such an attack now took place, however, and it was founded 

 upon a series of facts which I must relate here. In the bleach- 

 ing of wax by means of chlorine, Gay-Lussac had observed that 

 for every volume of hydrogen eliminated, an equal volume of 

 chlorine was taken up. 29 He had also found the same thing 

 in the action of chlorine- on hydrocyanic acid. In the course 

 of their investigation of the benzoyl compounds, previously 

 referred to, Wohler and Liebig, when acting with chlorine upon 

 bitter almond oil, had discovered benzoyl chloride ; and they 

 expressly remark that this substance is produced from the 

 bitter almond oil by two atoms of chlorine taking the place of 

 two of hydrogen. 30 In 1834, Dumas examines the action of 

 chlorine on oil of turpentine, 31 and in this case also each 

 volume of hydrogen eliminated is replaced by the same volume 

 of chlorine. Then when he studies the products of the de- 

 composition of alcohol by means of chlorine and of bleaching 

 powder, in order to clear up the nature, and the mode of 

 formation of chloral and of chloroform, he states the empirical 

 rule, observed in a single case by Gay-Lussac, in the following 

 general form : 32 



29 Gay-Lussac, Legons de Chimie. 30 Annalen. 3, 263. 31 Ann. Chim. 

 [2156,140. 32 Ibid. [2] 56, 113. 



