Theory of the Chemical Elements. 870 



These gases contain, throughout the series, a possibility of 

 disintegration into the gases of the atmosphere. 



Another vacancy precedes gold in the Table, the probable 

 atomic weight concerned being 197. We suggest the gas 



X 4 = 2{He 4 (P£H) 3 .4He 2 (P£H) 3 } 



with an atomic weight 196*76, exhibiting both the double 

 and quintuple groupings of the more primary groups. 



We have passed over a vacancy on the horizontal row con- 

 taining a sei'ies of the metals of the rare earths. There is a 

 current view that samarium '150*4), europium (152*0), and 

 gadolinium (157*3) form a triad in the last column of the 

 table, like iron, nickel, and cobalt, or osmium, iridium, 

 and platinum. Terbium (159*2) then appears in the column 

 with copper, silver, and gold, and should he preceded by an 

 inert gas of the neon group with an atomic weight nearly 

 equal to 158. If this view be accepted, the only compound 

 of the usual groupings which is available is, with atomic 

 weight 157-0', 



X 6 =4{He 4 (P£H) 3 . Nn 2 (PfH) 3 }. 



which is not quite in line with the general scheme. 



But a consideration of the table appears to indicate that 

 the true value should be nearly 153, if the successive differ- 

 ences are accounted in the usual way. The position of the 

 rare earths is a baffling problem, and there is certainly no 

 resemblance between terbium on the hand, and copper, 

 silver, and gold on the other. In fact, the rare earths all 

 resemble lanthanum, the first member of their series, in 

 properties, and there is considerable justification for the 

 proposal which has been made to extend the row to 16 

 instead of 8 members at this stage of the table. Quite apart 

 from the present theory, it does not seem that terbium can 

 retain its place, on purely chemical grounds, in the alkali 

 column. We shall suppose, therefore, that 153, the value 

 naturally suggested by the Periodic Table, is the proper 

 approximate atomic weight of the gas now in question. We 

 suggest therefore the formula 



X 5 =2{2He 2 (P£H) 3 . 3Nu 2 (PfH) 3 } 



with an atomic weight 152*49. There are other important 

 considerations in favour of this formula, which cannot be 

 discussed at present without undue digression. They will 

 be dealt with to some extent later in the paper, and more 

 specially afterwards, We may. however, state that there is 



