Theory of the Chemical Element*. 869 



is provisionally called protofiuorine, mainly for lack of a 

 better name. The name has been used before for a hypo- 

 thetical element with an atomic weight nearly the same. 

 No chemical similarity to fluorine is necessarily denoted. 

 This element appears to be present strongly in the solar 

 corona*. 



The investigations of the spectra of these elements are to 

 be communicated to the Royal Astronomical Society, and 

 will probably be published in part before this paper. 



The Nature of " Compounds " 



When an element is said subsequently to be a u compound " 

 of: simpler elements, the statement does not relate to a com- 

 pound as ordinarily known to chemists. Something of a 

 more intimate nature is indicated. On the modern view of 

 chemical affinity, valency is believed to be a capacity for 

 taking up or giving off a certain number of electrons, so 

 that when two elements, or ions, have opposite charges, they 

 may be held together in a compound. The " residual " 

 attraction of two molecules may also form an effective 

 compound, as in Sir J. J. Thomson's discussion of the 

 probable nature of chemical action |. But in ail truly che- 

 mical compounds, the separate atoms of the elements concerned 

 must be supposed to preserve their identity, though held 

 together by chemical forces, that is to say, electrical forces 

 of a particular kind. 



This is illustrated from the fact that compound gases have 

 a ratio of specific heats determinable from the number of 

 atoms in the molecule. Now the radium emanation or 

 niton belongs to a group of gases whose other members 

 are monatomic, and this gas is capable of giving off helium — 

 a fact which is difficult to explain if it be monatomic. The 

 difficulty can be evaded by supposing that the atom of the 

 gas contains all the components of the atom of helium, but 

 not as a compound, in any ordinary sense, of the atom of 

 helium with something else. In suitable circumstances, 

 these components are excluded together as a helium atom ; 

 but while in the atom of niton they are more intimately 

 related to the major part of the atom. Niton could therefore 



* The strongest lines in the coronal spectrum are due to this substance, 

 and it is therefore perhaps advisable to interchange the names coronium 

 and protofluorine ; but in this paper the names used when the paper was 

 read have been retained in order to avoid confusion. 



f Vide i The Corpuscular Theoiy of Matter.' 



Phil. Mag. S. 6. Vol. 22. No. 132 Dec. 1911. 3 M 



