DISCOVERT BY MEANS OF INFERENCE. 597 



vessel filled with air ; the bird died in forty-five minutes, 

 and he concluded that because a candle also became ex- 

 tinguished under similar circumstances there was a vital 

 fire in the body of the bird. 



Newton appears to have been the first to infer the 

 existence of chemical affinity ; he concluded that because 

 iron dissolves in a solution of cupric nitrate, and precipi- 

 tates the copper, that it has a stronger affinity than copper 

 for nitric acid. Other chemists also inferred the existence 

 of electric chemical attraction. ' That a body which is 

 united to another, for example, a solvent which has pene- 

 trated a metal, should quit it to go and unite itself with 

 another which we present to it, is a thing of which the 

 possibility had never been guessed by the most subtle 

 philosophers, and of which the explanation even now is 

 not easy.' Greoffroy, a talented French physician, said in 

 the year 1718 : ' We observe in chemistry certain rela- 

 tions amongst different bodies, which cause them to unite. 

 These relations have their degrees and their laws. We 

 observe their different degrees in this ; that among certain 

 matters jumbled together, which have a certain disposition 

 to unite, we find that one of these substances always unites 

 constantly with a certain other, preferably to all the rest,' 

 and, ' in all cases where two substances, which have any 

 disposition to combine, are united, if there approaches 

 then a third, which has more affinity with one of the two, 

 this one unites with the third and lets go the other.' 1 

 Bergmann, nearly one hundred years after Newton, from 

 the inference that different substances possessed different 

 degrees of chemical affinity, was led to discover and pub- 

 lish his series or tables of ' elective affinities,' in which 

 substances are arranged in the order of their different 



1 Whewell, History of the Inductive Sciences, 3rd edit. vol. iii. p. 100. 



