362 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP. 



Data are also available for antimony, bismuth and 

 tellurium, but no vapour-density determinations appear to 

 have been made in the case of magnesium, calcium, 

 strontium, barium, gold and platinum. It is clear that the 

 data are far too scanty to give trustworthy values for the 

 atomic weights of the metals. Thus the data for iron 



are : 



Molecular Weight of Iron in 

 Compound. Weight. Molecular Weight. 



Ferrous chloride 125*6 ... 56 (about) 



Ferric chloride 332 '2 ... 112 ,, 



Iron carbonyl 199*0 ... 56 ,, 



From so short a table we cannot tell whether 56, or a 

 sub-multiple of 56 (say 28), is the atomic weight of iron ; in 

 other words, we cannot be sure whether the molecule of 

 ferrous chloride or of iron carbonyl contains one, or two or 

 more atoms of iron. The facts show only that 56 is the 

 maximum value for the atomic weight of the metal. 



The formulae given in the last column of the large table are 

 based upon atomic weights deduced by other methods. They 

 show that the smallest weight of metal found in the molecules 

 of its volatile compounds is usually a single atom. But 

 pairs of atoms are frequently found 



(Al 2 Cl 6 ,Al 2 Br 6 ,Al 2 I G ,Fe,Cl c , etc.). 



In the case of copper the only compound examined is of 

 this type, the atomic weight of the element being 63 '6 and 

 not 127*1. 



B. THE LAW OF ATOMIC HEATS. 



Dulong and Petit's law of atomic heats (1819). The 



difficulty of finding the atomic weights of the metals by 

 means of density determinations renders very important the 

 fact, discovered in 1819 by the French physicists Dulong 

 and Petit (Ann. Chim. Phys., 1819, 10, 405), that 



" The atoms of all simple substances have the same capacity 

 for heat." 



