Influence of Atomic Weight. 309 



Now it has already been pointed out that the harder a body 

 the higher is its melting-point ; and as the hardness likewise 

 varies inversely as the specific volume, therefore the greater 

 the specific volume of an element the lower ought its melting- 

 point to be ; and this we actually find to be the case ivith the 

 elements taken as a whole. This will be again referred to 

 further on, in speaking of Meyer's " Curve of the Elements."" 

 As the tenacity of a metal (or the weight required to break a 

 rod of unit cross section) depends on the cohesion* between 

 the constituent particles of the metal, we should expect that 

 those metals which have the highest melting-points would also 

 be the most tenacious ; and this is really the case. Metals 

 like Fe and Cu, which melt at comparatively high tempera- 

 tures, have a far greater tenacity than such metals as Zn, PI, 

 Sn, which have but little tenacity. 



We have thus endeavoured to show how many of the physical 

 properties of the elements are interconnected with one another. 



One of the chief objects of the chemist and physicist of the 

 present day is to refer all the properties of the elements, both 

 chemical and physical, to as few, what we may call, standard 

 properties as possible, till finally one standard property is ob- 

 tained, to which all the others may be referred in some way 

 or other ; or, in other words, we have finally to choose some 

 standard property, of which all the others are a function, so 

 that when we are able to explain this standard property we 

 shall at the same time be able to arrive at the cause of the 

 other properties, and thus be in a condition to predict the 

 nature and degree of the properties of any given unknown 

 element, or any unknown properties of a known element, of 

 which the standard property has been determined numerically. 



Now, whenever possible, we should select as our standard 

 properties those which can be represented numerically, as the 

 atomic weight, specific gravity, melting-point, boiling-point, 

 &c; and then of these take as our final standard property that 

 which can be determined and represented numerically in the 

 most exact manner, and which is subject to least variation with 

 external circumstances. Thus, we might take as our ultimate 

 standard the coefficient of expansion, since this is capable of 

 pretty exact determination ; but it would not be advisable to 

 do so, as it is liable to great variation with the physical con- 

 dition of the body. The tendency at present (and it is no 

 doubt the right tendency) is to take the atomic weight as the 

 ultimate standard, and refer all the other properties to it ; for 

 it is capable in most cases of very exact determination, as Stas's 



* Mohr has recently shown that cohesion is hut one form of chemical 

 affinity, Liebig-'s^L?™. cxcvi. p. 183. 



Phil. Mag. S. 5. Vol. 8. No. 49. Oct. 1879. Y 



