POPULAR SCIENCE 287 



darkening, or quasi-continuous band, at the background of the 

 line and fluted spot spectrum. The appearances presented by 

 this complicated spectrum would seem to be for the most part 

 due to the absorption of a mass of metallic vapours and gases 

 of great density, the varying density or thickness of the strata 

 being denoted by the variations in the general darkening of 

 the background of the spot spectrum. 



The lines that are most widened in the sun-spot spectrum 

 are due to vanadium, titanium, and silicon, and it is generally the 

 fainter lines of their normal photospheric spectrum that are the 

 most affected by widening, an indication of the low level of a 

 sun-spot. Next in order of widening, at least in the region of 

 the red and yellow parts of the spectrum, which has been the 

 portion especially studied at Stonyhurst, come calcium, sodium, 

 nickel, manganese, iron, while cobalt, scandium, and chromium 

 are represented by well-marked lines. It is noteworthy that 

 most of the elements enumerated form the first half of the first 

 long period in the chemical periodic system. This would seem 

 to indicate that the level of sun-spots is that of the metals 

 with atomic weight about 50, for it is conceivable that, as we 

 advance outwards from the solar photosphere through a 

 mixture of metallic vapours and gases, there should be a layer 

 or stratum in which vapours of the metals connected by any 

 one period of the periodic law should be most abundant. 



Titanium also appears as titanium oxide in the banded 

 spectrum of sun-spots, in which, too, are represented mag- 

 nesium hydride and calcium hydride, with possibly a much more 

 familiar compound in the form of superheated steam. A 

 banded spectrum of chemical compounds is generally indicative 

 of a lowering of temperature. Hence it is argued that sun-spots 

 are at a lower temperature than that of the photosphere. To 

 the contrary, however, it is urged that a banded spectrum 

 may be produced by pressure alone, and, moreover, some com- 

 pounds, though titanium oxide is probably not among them, 

 are formed with an absorption and not with an evolution of heat. 

 If, however, appeal be made to the laboratories of the stars, 

 titanium oxide in particular, as Prof. Fowler has shown, is 

 responsible for a whole series of characteristic bands in the 

 spectra of stars of Type III., such as a Herculis. It is generally 

 supposed that Type III., or red stars, are at a lower temperature 

 than the yellow stars of Type II., which give a spectrum 



