NO. 1865 OUR knowledge; of the sun HALE 343 



ranged various light sources, in each of which the physical conditions 

 can be controlled by the observer. One of the simplest ways of vapor- 

 izing iron is to place fragments of the metal between the carbon poles 

 of an ordinary arc light. By varying the amount of metal present in 

 the arc, the effect of change of density of the vapor can be observed. 

 To study the influence of change of pressure, the arc must be in- 

 closed within a chamber, so constructed that air or some other gas 

 can be admitted and raised to the desired pressure. The effect is to 

 shift the lines of the spectrum toward the red, and by measuring 

 the displacement produced by an increase in pressure of one atmos- 

 phere, the pressure within a Sun-spot or in a star, corresponding to 

 any observed shift of the lines, can be determined. To ascertain 

 the effect of change of temperature upon the spectrum, the iron 

 vapor at the very hot center of the arc may be compared with the 

 cooler vapor in the outer part of the flame. If the highest tempera- 

 ture of the arc is not sufficiently great, a powerful electric spark, 

 taken between two poles of iron, will afford a still hotter light- 

 source. Apparatus suitable for all of these purposes and for other 

 similar ones is arranged upon the annular pier. When the light from 

 any particular source is to be investigated, it is reflected from a 

 plane mirror at the center of the circle to a concave mirror (shown 

 near the middle of Plate xxix), which forms an image of the source 

 on the slit of a powerful spectrograph. 



For various reasons it seemed probable that reduced temperature 

 might be the cause of the strengthening and weakening of lines in 

 spot spectra. Accordingly, special attention was directed to a study 

 of the effect of temperature change on the relative intensities of the 

 lines. After an extensive investigation it was found that the iron 

 lines whose relative intensities increase at reduced temperatures are 

 invariably among the lines which are strengthened in Sun-spots. 

 Moreover it was also found that the iron lines which are weakened 

 at reduced temperatures are weakened in Sun-spots. After these 

 experiments had been extended from iron to titanium, vanadium, 

 chromium, manganese, cobalt, nickel, and other substances con- 

 spicuously represented in Sun-spots, the conclusion was reached that 

 a reduction in temperature of the spot vapors is competent to ex- 

 plain a large part of the characteristic spectral phenomena. Assum- 

 ing this hypothesis to be correct, one would naturally be led to ask 

 Avhether the temperature of the spot vapors is sufficiently reduced to 

 permit elements existing uncombined at the higher temperature of 

 the solar surface to enter into combination within the spot. Titanium 

 and oxygen, for example, both occur among the vapors which lie 



