KLAPROTH AND DALTON 177 



that our ideas have approached considerably nearer to actual 

 reality. Furthermore, Count Rumford had directly proved 

 that there can be no question of heat having weight.^ 



From this point, Dalton again made a great advance. He 

 saw clearly the great importance of the fixed relationships by 

 weight with which any two components combine to form a 

 new substance: the 'law of constant proportion' as it is 

 called. To this was added the 'law of multiple proportions,' 

 inasmuch as it was found that in many cases several different 

 relationships by weight of the same components were possi- 

 ble, but that these were then multiples of one another, and 

 also, these multiples were always small whole numbers. 

 Dalton had himself specially examined some examples of 

 this. 



Along with these facts, Dalton had in mind the extreme 

 divisibility of all bodies; not only of simple bodies, but also of 

 compound ones, such as water, which are not mixtures capa- 

 ble of variation in composition. Since division, no rriatter 

 how far it is carried, leaves their properties unchanged, the 

 smallest parts must still contain the components, for ex- 

 ample hydrogen and oxygen, in the correct and fixed rela- 

 tionship by weight in which they are always found in any 

 quantity whatsoever of the same substance. He concluded 

 from this that such bodies as water (we call them to-day 

 chemical compounds, as contrasted with mixtures) must 

 consist finally of very small parts, which are all exactly alike, 

 and which are themselves again composed of still smaller 

 parts of the given fundamental substances, which latter, how- 

 ever, in spite of their obviously imperceptible minuteness, 

 must still have the quite definite weight corresponding to the 



1 In the matter of heat, Count Rumford and Lavoisier agree. The 

 difference is however, that the practical imponderability of heat was in 

 Rumford 's case a fact of observation proved conscientiously, whereas in 

 Lavoisier's case it was a hypothesis introduced a priori; further that 

 Rumford - again as the result of direct experiment - did not regard 

 quantities of heat as invariable in all circumstances, whereas Lavoisier 

 introduced his 'caloric,' which was supposed to be invariable in amount. 



Ns 



