MEDICAMENTS AND LIFE. 



is inverse for those which are related to chlorine, and for 

 those of the arsenical class. This untiring investigator, in 

 order to establish these laws, undertook experiments six 

 years ago, which have been steadily followed up till very 

 lately ; and the Academy of Sciences has recognized their 

 discovery by a brilliant reward. Their practical interest 

 may easily be estimated. When a physician, in future, 

 has need to choose between different salts, all that will be 

 required will be to consult a table of atomic weights in 

 order to learn at once their respective activities, and con- 

 sequently to fix upon the proper dose of them. When a 

 physiologist wishes to test the action of a metallic com- 

 pound, he will be able to foretell its relative intensity, and 

 thus to guide his use of it in experiment. When the effect 

 of salts of thallium was tried upon animals some years ago, 

 this being one of the metals just revealed by spectrum 

 analysis, it was noted with much surprise that these salts, 

 so extremely similar in other respects to those of soda and 

 of potassa, were yet powerfully poisonous. This is because 

 the atomic weight of thallium is very high ; its poison- 

 ing potency is thus in exact agreement with Rabuteau's 

 law. 



The improvement of the healing art is thus allied in 

 the closest way with the advance of our knowledge as to 

 the true action of toxic and medicinal substances. To 

 enlarge this knowledge, we must follow Bernard's ex- 

 ample and methods in the examination of effects pro- 

 duced upon animal tissues. It is of moment, too, as Du- 

 mas recommends, to test the action of all those new sub- 

 stances which organic chemistry has been for some time 

 producing, several of which undoubtedly carry in them 

 medicinal qualities. The study of these effects is very- re- 

 fined, and those savants who undertake it will need to 

 handle with equal skill the instruments of physics, of physi- 

 ology, and of chemistry. The point is not simply to analyze 



