STANISLAO CANNIZZARO 155 



Cannizzaro then gives a table of molecular weights, referred, 

 first to the weight of a molecule of hydrogen as unity, and 

 secondly, to the weight of a semi-molecule of hydrogen as unity; 

 in other words, he tabulates the densities of certain elementary 

 and compound gases, referred, first to h3^drogen as unity, and 

 secondly, to hydrogen taken as having a density equal to two. 



Cannizzaro's table contains two values for the density, and 

 therefore, for the molecular weight, of oxygen, and two values 

 for sulphur-gas. The value given to the molecular weight of 

 oxygen, referred to that of hydrogen as twice unity, is 32 ; and 

 the value for "electrified oxygen" is 128. To the molecular 

 weight of sulphur "under 1000''" is given the value 192, and 

 to that of sulphur " over 1000"," the value 64. At the very 

 outset of his research Cannizzaro recognised the coalescence 

 into groups, which act as individual molecules, of different 

 numbers of identical elementary atoms. He brought within the 

 scope of the molecular and atomic theory some of the phenomena 

 of allotropy ; and, so far as his application of the theory went, 

 he made it possible to think of these phenomena without using 

 the shuffling phrase, " different forms of an element." 



Cannizzaro's next step was to examine the compositions of 

 molecules the relative weights of which he determined by using 

 the hypothesis of Avogadro. This is the part of his work 

 which seems to me most evidently marked with the sign-manual 

 of a splendid clarity. He describes his method with admirable 

 precision : 



" If the body is a compound, it is analysed, and the constant 

 weight-relations of its constituents are determined ; the mole- 

 cular weight is then divided into parts proportional to the 

 relative weights of the components, and the result is the quan- 

 tities of the elements contained in the molecule of the compound, 

 referred to the same unit [namely, the semi-molecule of hydro- 

 gen] as is used for the expression of all molecular weights." 



In applying this method Cannizzaro first sets down the 

 molecular weights of various gases, both elementary and com- 

 pound, " referred to the weight of a semi-molecule of hydrogen 

 as unity," and then the weight (again referred to that of a 

 semi-molecule of hydrogen as unity) of each constituent of each 

 gas contained in a molecular weight of that gas. He then 

 makes similar arrangements of the analyses of several com- 



