so-called Meta- Elements, 119 



this element in the table if its atomic weight be taken as this 

 number. If, however, it be considered a tetravalent element, 

 its atomic weight becomes 165, and it fits in well in the 

 fourth group of the ninth series. It is very probable that 

 terbium should be a tetravalent element forming salts of the 

 formula Tb 2 X 4 ; for cerium, the commonly occurring -oxide of 

 which is Ce 2 3 , is well known to form these salts, and terbium 

 is very similar in properties to this element. 



The next element, erbium (166), finds its place in the fifth 

 group of this series, whilst the next, ytterbium (173" 7), is 

 placed beneath lanthanum and holmium in the third group 

 of the tenth series. The only element in the list for which 

 a vacant place does not appear is thulium (170*7) ; very little 

 if anything is known of the properties of this element, and so 

 it is scarcely possible to say which group it had better be 

 placed in : it does not appear as if it would fit any of the 

 places in the table at present empty, and so, if further re- 

 search tends to confirm its being a trivalent element with 

 this atomic weight, it will certainly constitute a difficulty not 

 to be lightly got over. 



We have thus found, with perhaps one doubtful exception, 

 that there are places in the table of elements for all the rare 

 earths discovered up to the present time, and, moreover, that 

 these places agree well with such properties of these elements 

 as are at present known. It may also be noticed that there 

 are several empty places ready to receive any more newly- 

 discovered elements. 



The conclusion which may be drawn from the results 

 arrived at in this paper are, that it would seem quite un- 

 necessary to suppose the actual decomposition of what was 

 previously supposed to be an element takes place when an 

 earth is fractionated in some suitable manner as by partial 

 precipitation. If such a decomposition is possible for one or 

 two elements, we may reasonably expect it for all. Why, 

 then, is it not attempted to fractionally separate an element 

 which can be obtained in a pure state to start with ? In the 

 fractionation of these earths, it is almost if not quite im- 

 possible to obtain the earth in a pure state in the first 

 instance ; that is, to obtain an earth which is known not to 

 contain at least a small quantity of some other earth totally 

 foreign to it (disregarding the several so-called meta- 

 elements into which it is to be separated). The difficulty 

 is also enormously complicated by the fact of the earths 

 which are attempted to be fractionally separated having 

 almost exactly the same properties as at least half a dozen 

 other earths. 



