20 



Prof. T. Carnelley on the Periodic Law. 



those calculated limits (see Table IX.) for little- known or undis- 

 covered elements within which these lie, and we should obtain 

 as follows: — 



Chlorides. 



M.P. of 

 Titanium 

 chloride 

 not de- 

 termined. 



B.P. found for 

 Titanium 

 chloride 



=408°. 



B.P. calculated 

 for the posi- 

 tion usually 

 assigned to 

 TiCl 4 



= (405-408°). 



Ditto for posi- 

 tion usually 

 assigned to 

 Yt01 3 



= (388-421°) 



Bromides. 



M.P. found for 



Titanium 



bromide 



=312°. 



M.P. calculated 

 for position 

 usually as- 

 signed to 

 TiBr 4 



= (295-343°). 



Ditto for posi- 

 tion usually 

 assigned to 

 YtOL 



= (329-351°). 



B.P. found for 

 Titanium 

 bromide 



= 503°. 



B.P. calculated 

 for position 

 usually as- 

 signed to 

 TiBr 4 



= (493-507°). 



Iodides. 



M.P. found for 



Titanium 



iodide 



=423°. 



M.P. calculated 

 for position 

 usually as- 

 signed to 

 Til 4 



= (414-442°). 



Ditto for posi- 

 tion usually 

 assigned to 

 Yt01 3 

 = (419-428°). 



Ditto for posi- 

 tion usually 

 assigned to 

 EkaSnI 4 



= (428-490°) 



B.P. found for 



Titanium 



iodide 



= 633°. 



B.P. calculated 

 for position 

 usually as- 

 signed to 

 Til 4 

 above 614°. 



Ditto for posi 

 tion usuall 

 assigned to 

 EkaSnI 4 

 = (568-654 c 



This table shows that Ti must occupy in the classification 

 either the position assigned by Mendeljeff to the little inves- 

 tigated Yt (to the exclusion of this latter element), or that of 

 the undiscovered element EkaSn, or, lastly, that which it 

 really occupies ; but it cannot belong to either of the first two 

 positions, for neither do the experimental boiling-points of its 

 bromide and iodide agree with those calculated for the first 

 position, nor do the experimental melting-points and boiling- 

 points of its chloride and bromide agree with those calculated 

 for the second position; nor, again, does its equivalent agree 

 with either of those which the first and second positions 

 require. Consequently the only position which remains is that 

 usually assigned to Ti; and this also agrees with its experi- 

 mental equivalent. 



Boisbaudran has found that gallium chloride melts at 348 o, 0. 

 Taking this melting-point and the equivalent 23*3 as our only 

 available data, and proceeding as described above in the case 

 of Ti, we find that we must give to gallium the position of 

 Eka-aluminium in Mendeljeff's table. This is the position 

 actually assigned to it for other reasons hy Boisbaudran. 

 There has recently been some dispute as to the atomic 



