Influence of Atomic Weight. 323 



In divisions 2 and 3 the elements on the falling portions of 

 the curve are electro-positive and electro-negative > n the 

 rising parts, whilst in divisions 4, 5, and 6 those elements 

 following directly after the maxima and minima are positive, 

 and those occurring immediately before these points are nega- 

 tive. A well-marked positive or negative character is possessed 

 only by those metals which have comparatively large atomic 

 volumes, or, according to Meyer, " the aggregation of a large 

 mass in a small space appears to be incompatible with the 

 development of a marked positive or negative character." 



On comparing the Table of melting-points (p. 315) with 

 Meyer's curve of the elements, we find that the melting- 

 points (and also the boiling-points) rise and fall as the curve 

 falls and rises; i. e. they are inversely as the atomic volumes*. 

 The only important exceptions to this rule are As and Se in 

 series 5, Sn, Sb, and Te in series 7, and Tl, Pb, and Bi in 

 series 11, the melting-points of which are too high to be in 

 accordance with the above rule. I have pointed out (JDeut. 

 chem. Ges. Ber. xii. 440), as previously mentioned, that the 

 coefficients of expansion of the elements are the greater the 

 lower their melting-points, and that there are but five excep- 

 tions to this rule, viz. As, Sb, Bi, Te, and Sn. Now it is seen 

 that all these five elements are also among the eight excep- 

 tions to the relation between the melting-point and atomic 

 volume, as is shown in following the course of Meyer's curve. 

 All these eight exceptions likewise belong to odd series and 

 follow one another directly in their respective series. A fur- 

 ther inspection of Meyer's curve shows that it rises regularly 

 and somewhat sharply from Ni, Cu, and Zn to Ga, and then, 

 though it does not really do so, yet it shows a very strong ten- 

 dency to turn downwards to As. In the same way, in division 

 5 the curve rises regularly and rapidly from Pd, Ag, and Cd 

 to In, and then exhibits a strong downward tendency towards 

 Sn and Sb. These facts may perhaps serve to explain why the 

 melting-points of the elements As, Se and Sn, Sb, Te at 

 these points on the curve are not regular either in relation to 

 the atomic volume or to their coefficients of expansion. The 

 same thing applies also in the case of Pb. Therefore, though 

 at first sight these exceptions appear to contradict, yet a closer 

 inspection shows that they rather confirm the course of Meyer's 



* This is only what we might expect from the remarks on p. 308 with 

 regard to JBettone's experiments. For since, according to him, the hard- 

 ness of an element = : =-, and since the harder a body, the higher 



atomic vol.' J ' G 



its melting-point, then the greater the atomic vol. of an element, the more 

 easily fusible it is. 



