310 Dr. T. Carnelley on the 



classical researches have shown, and for a given element is, as 

 far as we know, absolutely invariable. 



This subject (viz. the reference of the properties of the ele- 

 ments to their atomic weights) was first attacked in a syste- 

 matic manner by Newlands (" Law of Octaves," Chem. News, 

 xiii. p. 113, also ibid. x. pp. 59, 194), who in 1864 pointed 

 out that the atomic weights and properties of the elements 

 vary periodically with one another ; and in 1869 Mendeljeff, 

 apparently independently of Newlands, in an elaborate and 

 most important paper, propounded what is known as the 

 Periodic Law (JJeut. chem. Ges. Ber. ii. p. 553, Ann. Chem. 

 Pliarm. Suppl. viii. pp. 133-229). This law states that 

 " The properties of the elements are a periodic function of their 

 atomic weights" What Mendeljeff in reality did was to take the 

 atomic weight as his final standard, and represent each of the 

 other properties as a function of this standard. MendeljefF's 

 researches refer chiefly to the relations between the atomic 

 weights and properties of the elements ; and before going on 

 to the question of compounds, we shall direct attention to some 

 of the more salient features of his work. He points out that 

 if the elements be arranged according to the size of their 

 atomic weights from H = 1 to U = 240, then the relations be- 

 tween their atomic weights and their chemical and physical 

 properties exhibit a periodic function. As an example, take 

 those elements the atomic weights of which lie between 7 and 

 36, thus : — 



Li =7 Be = 9'2 B =11 C =12 N = 14 = 16 F =19 

 Na=23 Mg = 24 Al = 274 Si=28 P=31 S =32 Cl = 35-5 



Here the elements up to F are arranged in the first line and 

 those after F in the second line, each element as it comes 

 being placed immediately under the one above. It is readily 

 seen that those elements which stand in the same vertical 

 column have very similar properties, and are in fact always 

 classed as belonging to the same family or group. It is there- 

 fore evident that the character of these elements changes 

 regularly and gradually with the increase in their atomic 

 weights ; and that this variation is a periodic one, i. e. varies 

 in the two series in a similar manner. As a further example 

 of this, take the composition of the oxides of the above four- 

 teen elements, and it will be seen that the corresponding 

 members of both series form oxides having a similar composi- 



(ozone) F 2 O (?) 

 C1 2 



This shows also that those elements in the same vertical 

 column have the same atomicity, and that the atomicity in- 

 creases regularly up to the middle member of each series and 



tion ;. thus : — 











Li 2 BeO 



B,0 3 



C0 2 



N 2 3 



00. 



Na 2 MgO 



A1 2 3 



Si0 2 



p,o 3 



so 2 



