Mr. M. M. Pattison Muir 07i Chenmal Classification, 267 



which phosphorus does on the ascending. Arsenic and antimony 

 hold positions respectively in the fourth and fifth sections of the 

 curve, somewhat similar to those occupied by phosphorus and 

 nitroo-en. But the fourth and fifth sections, as has been already 

 pointed out, comprise each two minor sections within itself. 

 Arsenic follows immediately upon a number of well-marked 

 positive elements (cobalt, nickel, copper, and zinc). The 

 influence of these appears to be exercised upon the negative 

 arsenic. This is even more marked in the case of antimony. 

 This element is immediately preceded by a long list of positive 

 elements (rhodium, palladium, silver, cadmium, and indium), 

 from which it is only separated by tin : it is followed by 

 iodine. Bismuth is even more hemmed in, so to speak, by 

 positive metals ; the difference between its atomic volume and 

 the vokunes of its immediate predecessors in the curve is not 

 great; and, so far as we know, it has no electro-negative 

 element, such as fluorine, chlorine, bromine, or iodine (which 

 respectively succeed nitrogen, phosphorus, arsenic, and anti- 

 mony), coming after it. The absolute increase in atomic 

 weight as we ascend from nitrogen to bismuth, also, it would 

 appear, influences the power of these elements to combine 

 with oxygen and to form stable oxides. The greater or less 

 fusibility of the elements themselves and of their immediate 

 neighbours must doutless also be considered. Nitrogen is 

 surrounded by gaseous elements, phosphorus by those which 

 have low melting-points, arsenic by comparatively easily 

 fusible elements, antimony by elements which are more fixed, 

 and bismuth by elements of somewhat high melting-points. 



Now if we turn to vanadium, niobium, and tantalum, we 

 find that they are situated on descending portions of the curve, 

 that, of the three, vanadium is nearest a minimum-point and 

 nearest also to a number of positive elements. There is a 

 very considerable distance between either niobium or tan- 

 talum and a positive element ; and these metals are further 

 protected on each side by the interposition of negative 

 elements. 



Considering these points (and others might be adduced), 

 there is good ground for believing that the behaviour of the 

 eight elements constituting the nitrogen group, towards 

 oxygen and chlorine respectively, is closely connected, as a 

 periodic function, with the atomic weights of these elements. 



The position of the elements on the periodic curve is seen 

 to be connected with the valency of these elements as deduced 

 from a study of their gaseous chlorides, hydrides, ethides, &c. 

 Thus the second and third sections of the curve are con- 



