THE ELEMENTS OF THE CHEMIST 133 



leeff's ekaboron and the metal scandium when examined by 

 Cleve a few years later. 



One other application of the periodic law may be mentioned, 

 and that is the guidance it has afforded in correcting the atomic 

 weights. The element glucinum had formerly the value 13*5 

 assigned to it. There is, however, no place in the periodic 

 scheme for an element of this atomic weight with properties 

 such as those exhibited by glucinum. A further investigation of 

 its properties and determinations of its specific heat showed 

 that the atomic weight was much lower, and the figure 9-1 

 entitles it to a place in the table in Group II, next above mag- 

 nesium, with which it has considerable analogy. Other cases of a 

 similar kind have led to correction which all experience tends to 

 verify. Two elements only present outstanding difficulties. 

 These are the elements argon and tellurium, both of which are 

 placed in the list of elements one step too high in the consecu- 

 tive order, notwithstanding all the very numerous experimental 

 investigations as to the numerical values of their atomic weights. 

 So strong is the general conviction that their true places in the 

 Mendeleeff scheme are those which have actually been assigned 

 to them, notwithstanding the numerical discrepancy. 



Of course it must also be admitted that in Mendeleeff's table 

 satisfactory places have not as yet been found for some of the 

 elements derived from the rare earth minerals, so that the 

 cerium and yttrium and other groups do not quite fall into line. 



The element argon has been mentioned, and we must now 

 review as briefly as may be the dramatic story of its discovery. 



Previously to 1894 the existence of a group of elements 

 destitute of all power to enter into chemical combination had 

 not been foreseen by Mendeleeff or any of the chemists who had 

 for years made a study of the periodic scheme. But for several 

 years Lord Rayleigh had been engaged in a series of experiments, 

 the object of which was to determine with the utmost possible 

 accuracy, not only the relative densities, but the absolute 

 densities, of the principal gases, that is, to compare their weights 

 with that of an equal bulk of water. The method used was the 

 same as that which had been employed by the French physicist 

 Regnault many years before, and is in principle of the utmost 

 simplicity. It consists in weighing the gas contained in a large 

 glass globe attached to one arm of a balance, using as a counter- 

 poise a similar globe of as nearly as possible the same size, so as 



