On the Temperature of Sun-spots. 303 



in the solid or liquid state, emitting a continuous spectrum ; 

 and that by some means it is further heated until it becomes 

 gaseous, as we have it in the electric arc ; it will now emit 

 very little visible radiation but some very strong ultra-violet 

 rays. The radiation from silicon in the visible part of the 

 spectrum appears to be not relatively but absolutely less when 

 it is gaseous than when it is solid. The whole argument of 

 the gradual shortening of the period of oscillation with increase 

 of temperature hangs on the supposition that there is a con- 

 tinuity of state, which must not be assumed to be the case in 

 sun-spots. The phenomena of spots are those of increased 

 absorption by (1) increase in the number of Fraunhofer lines, 

 and (2) increase in their breadth or in their darkness. Now 

 the increase in the number of Fraunhofer lines may be due 

 either to an addition to the sun's atmosphere of the vapours of 

 some substances not usually present, or to an increase in 

 the rays absorbed by those which are usually present. 



Increase of temperature frequently causes vapours to emit 

 (and absorb) rays not seen at lower temperatures, and increase 

 of temperature might add to the number of different gaseous 

 substances in the sun's atmosphere by evaporating materials 

 which had been before solid or liquid. On the other hand, 

 cooling might allow of the formation of chemical compounds 

 with absorptive powers quite different from those of their 

 (separate) components. We cannot then safely infer either 

 an increase or a decrease of temperature from an increase in 

 the number of Fraunhofer lines. But, turning to the in- 

 creased width of the lines, we know no case in which such an 

 effect is produced by a diminution of temperature. Many 

 examples seem to indicate that such widening is connected 

 with increased density of the vapour producing the line ; not 

 however increased density produced by cooling, but only 

 increased density produced by increased evaporation, in- 

 creased (partial) pressure of the vapour. In fact, when a line 

 is self-reversed (as it has been called) the dark line due to the 

 cooler vapour is always narrower than the bright line emitted 

 by the hotter vapour, even when there is no reason to think 

 the hotter vapour to be more dense than the cooler, as in the 

 case of the reversal of the F line of hydrogen when sparks 

 are passed through hydrogen, as observed by Hasselberg 

 (" Ueber die Spectra der Cometen/' Mdm. de I Acad. Imp. St. 

 Petersbourg xxviii. No. 2, p. 74 note) and by ourselves (Proc. 

 Roy. Soc. xxx. p. 74). 



Any compression produced by a carrying downwards of 

 the atmosphere must be attended with an increase of temper- 

 ature. We have been driven by considerations of this kind 



