Celestial Chemistry. 395 



The lines beneath the spectra indicate the position of the prin- 

 cipal metallic bands, showing the coincidence or otherwise. 

 The colours of the spectra are the same as those of the solar 

 spectrum, but much fainter, Aldebaran being the stronger of 

 the two, and many of the extreme rays at both ends are too 

 faint for perception. 



Aldebaran is of a pale-red tint, and the spectrum is fall of 

 lines in the orange, green, and blue. We shall presently see 

 the significance of this fact. About seventy lines were measured, 

 and these being compared with the chart of the metallic ele- 

 ments, a number of approximate coincidences were selected 

 to be tested by a direct comparison of the stellar and metallic 

 spectra. Sixteen terrestrial elements were thus made to pre- 

 sent their characteristic bright lines above the dark ones of 

 the star, and sodium, magnesium, hydrogen, calcium, iron, 

 bismuth, tellurium, antimony, and mercury, answered perfectly 

 to the test, and proved their existence in the atmosphere of 

 Aldebaran. Seven elements, nitrogen, cobalt, tin, lead, cad- 

 mium, lithium, and barium, gave negative evidence only. 

 Indications of other elements were not wanting, but the obser- 

 vations of these fine lines were too fatiguing to the eye to 

 be pressed further. 



JBetelgeux (a Orionis) has a very full spectrum. The light 

 of the star is orange, and the lines are thickest in the red, 

 green, and blue. About eighty lines have been measured and 

 mapped, and sixteen elementary substances tested for coinci- 

 dence. With five of these, sodium, magnesium, calcium, iron, 

 and bismuth, the connection is complete, and certain. Per- 

 haps thallium also exists, but the line seen might be calcium, 

 the apparatus not having dispersive power enough to separate 

 the lines of the two metals at that particular point. In the 

 case of hydrogen, nitrogen, gold, cadmium, silver, mercury, 

 barium, and lithium, coincidence was not proved. By reference 

 to the coloured plate, the coincidences will at once be seen, and 

 the relative positions of the lines of the other elements which 

 were compared. 



/3 Pegasi gives a much fainter spectrum, and fewer lines 

 have been measured ; of this star a diagram is not given. 

 Sodium and magnesium, and perhaps barium, exist in its enve- 

 lope. Iron and manganese are doubtful, but nitrogen, tin, 

 mercury, and hydrogen were not present. The absence of 

 hydrogen from a Orionis and /3 Pegasi is a point of considerable 

 importance, as its presence is eminently characteristic of the 

 spectra of the sun and some forty fixed stars, which have been 

 examined. 



This observation is valuable, since it proves that the lines 

 which indicate its presence belong to atmospheres of the lumi- 



