540 



Sir N. Lockyer and Mr, F. E. Baxandall. [Jan. 3, 



the elimination of those due to other metals. There can be little 

 doubt that the majority of the lines, except, perhaps, some of the very 

 lowest intensity, really belong to scandium. Exner and Haschek's 

 wave-lengths and intensities of the scandium arc lines are given for 

 comparison. 



Scandium, Lines in the Solar Spectrum. 



Rowland, in his " Tables of Solar Wave-lengths," ascribes a small 

 number of lines to scandium, but a comparison of the Kensington 

 photographs of the arc spectrum of this element with the solar spectrum 

 shows that in addition to these there are other solar lines nearly 

 certainly due to the same element. The table gives the solar lines 

 which, by a careful comparison of the metallic and solar spectra, have 

 been considered to correspond, without any doubt, with scandium 

 lines. In addition to these, there is a considerable number which 

 agree closely in position with weak solar lines, but of their identity 

 there is, perhaps, some doubt. In some cases the solar lines are so 

 weak that it is impossible to establish their identity with scandium 

 lines by direct comparison of the two spectra, the only guide being the 

 close agreement in wave-length, and the relative intensity of the 

 metallic and solar lines. In other cases it is doubtful whether the 

 metallic lines are strong enough to account for the solar lines. In the 

 table these lines are denoted by an asterisk, and must be accepted only 

 provisionally as ' ' possible " scandium-solar lines. 



The following analysis of the scandium lines, with reference to their 

 intensities, and their appearance or non-appearance in the solar 

 spectrum, will be of interest. 



It will be seen that of the 23 lines of intensity 6 or greater, 18 

 occur in the solar spectrum, three others are doubtfully present, while 



