contributions to the Spectral Analysis of the Stars. 3G5 



would be found of separating all the lines which result from ab- 

 sorption in the earth's atmosphere from those which owe their 

 origin to the sun's atmosphere, inasmuch as the displacement in 

 question can only affect the latter. 



Another subject of investigation by spectrum-analysis of the 

 sun are the protuberances. Lockyer and Janssen have, as is 

 well known, succeeded in observing the spectrum of these ob- 

 jects (consisting of three bright lines) independently of a total 

 solar eclipse. 



At present attention is directed on all sides to finding out 

 methods which shall enable not only those lines, but the entire 

 figure of the protuberances to be simultaneously observed. 



The position of the bright lines corresponds to the magnitude 

 of the dimension of the protuberance in question which falls in 

 the direction of the slit. When the slit is brought successively 

 into various directions so that it cuts the protuberance in just so 

 many positions, we are in a position to construct the shape of 

 the body observed, as Lockyer has already done. Janssen has 

 proposed the construction of a rotating spectroscope, so that, with 

 adequate velocity of rotation, by means of the duration of the 

 impression of light the shape of the entire protuberance might 

 be seen at once. 



Apart from the mechanical difficulties of such a rotating spec- 

 troscope, in which one of the three bright protuberance-lines 

 must be exactly in the axis of rotation, the object in view might 

 be more simply and completely obtained by oscillating the slit 

 at right angles to its direction. We should then be in a position 

 to observe the same protuberance simultaneously in three differ- 

 ently coloured images corresponding to the three different lines of 

 its spectrum. Yet in these methods with a moveable slit, the 

 difference in brightness, through which the protuberance stands 

 out against the ground, is considerably enfeebled according to 

 the distance traversed by the slit. With the rotating spectro- 

 scope more especially, the brightness of the protuberance would 

 be weakened from the centre of rotation towards the edge, and 

 the observation of the natural relative brightness of the image 

 would be prevented. 



For this reason I intend using another very simple means for 

 attaining the object in question, of the practicability of which I 

 have convinced myself by experiments (to be subsequently de- 

 scribed) on terrestrial sources of light. The principles upon 

 which this method depends are the following: — 



(1) The apparent brightness (lustre, claritas visa*) of a protu- 

 berance-band is independent of the breadth of the slit, provided 

 that it always retains a perceptible breadth upon the retina. 

 * Lambert, Photometria &c. §§ 36 & 37. 



