300 A CENTURY OF SCIENCE 



ratio 1 to 16 for the weights of the hydrogen and oxy- 

 gen atoms, while with HO the ratio is 1 to 8. 



Berzelius attempted to bring about greater uniformity 

 in formulas and atomic weights by making changes in his 

 table of atomic weights published in 1826. He prac- 

 tically doubled the relative atomic weights of hydrogen, 

 chlorine, nitrogen, and of the other elements that gave 

 twice as many atoms in his formulas as in those of others, 

 and at the same time he wrote the symbols of these 

 elements with a bar across them to indicate that they 

 represented double atoms. For example, he wrote : 



HO Znl NO 6 

 instead of 



H 3 0, ZnCl, N a O 



This appears to have been an unfortunate concession 

 to the views of others on the part of Berzelius, for the 

 barred symbols were not generally adopted, partly on 

 account of difficulties in printing, and the great achieve- 

 ment in theory made by him was lost sight of for a long 

 period of time. , 



The Law of Atomic Heats. In 1819, Dulong and Petit 

 of France, from experiments upon the specific heats of a 

 number of solid elementary substances, came to the con- 

 clusion that the atoms of simple substances have equal 

 capacities for heat, or in other words, that the specific 

 heats of elements multiplied by their atomic weights give 

 a constant called the atomic heat. For instance, the 

 specific heats of sulphur, iron, and gold have been given 

 as 0-2026, 0-110, and 0-0324, while their atomic weights 

 are about 32, 56, and 197, respectively; hence the atomic 

 heats obtained by multiplication are 6-483, 6-116, and 

 6-383. 



Further investigations showed that the atomic heats 

 display a considerable variation. Those of carbon, 

 boron, beryllium, and silicon are very low at ordinary 

 temperatures, although they increase and approach the 

 usual values at higher temperatures. More recent work 

 has shown, however, that the specific heats of other ele- 

 ments vary greatly with the temperature, almost disap- 

 pearing at the temperature of liquid hydrogen, and hence 

 possibly disappearing entirely at the absolute zero, where 



