CHEMICAL AFFINITY. 165 



stance there would be no subdivision of the (supposed indivi- 

 sible) molecule ; and this held good with many compounds : 

 thus fourteen parts by weight, say grains of nitrogen, will 

 combine respectively with eight, sixteen, twenty-four, thirty- 

 two, and forty parts by weight, or grains, of oxygen. 



So, again, twenty-seven grains of iron will combine with 

 eight grains of oxygen or with twenty-four grains, i. e. three 

 proportionals of oxygen. No compound is known in which 

 twenty-seven grains of iron will combine with two propor- 

 tionals or sixteen grains of oxygen ; but this does not much 

 affect the theory, as such a compound may be yet discovered, 

 or there may be reasons at present unknown why it cannot be 

 formed. 



But now comes a difficulty : twenty-seven parts by weight 

 of iron will combine with twelve parts by weight of oxygen, 

 and twenty-seven parts of iron will also combine with ten 

 and two-third parts of oxygen. Thus if we retain the unit 

 of iron we must subdivide the unit of oxygen, or if we retain 

 the unit of oxygen we must subdivide the unit of iron, or we 

 must subdivide both by a different divisor. What then be- 

 comes of the notion of an atom or molecule physically indi- 

 visible ? 



If iron were the only substance to which this difficulty 

 applied, it might be viewed as an unexplained exception, or 

 as a mixture of two oxides ; or recourse might be had to a 

 more minute subdivision to form the units or equivalents of 

 other substances ; but numerous other substances fall under 

 a. similar category ; and in organic combinations, to preserve 

 the atomic nomenclature we must apply a separate multiplier 

 or divisor to far the greater number of the elementary con- 

 stituents, i. e. we must divide that which is, ex hypothesi, indi- 

 visible. 



Thus, to take a more complex substance than any formed 

 by the combination of iron and oxygen, let us select the sub- 

 stance albumen, composed of carbon, hydrogen, nitrogen, 



