132 KUMFORD SPEOTROHELIOGRAPH. 



The Avidespread interest in total solar eclipses and the great ex- 

 penditure of time and money so freely made in obserAdng them 

 surely tend to emphasize Avhat has been said. For if it is worth 

 while, as it certainly is, to travel thousands of miles and to undergo 

 hardships in order to spend a few flying seconds in making observa- 

 tions, it Avould seem no less advantageous to continue solar work at 

 home, Avhere entirely new phenomena can be observed daily with 

 a much smaller expenditure of effort. Total eclipses of the sun will 

 always be of great importance, as the corona can not l)e observed in 

 full sunlight. But the study of many other solar phenomena, which 

 can be observed whenever the sky is clear, is quite as likely to 

 advance our knowledge of the solar constitution. 



It was with some such ideas in mind that the work of the Kenwood 

 Observatory was undertaken in 1888. It seemed obvious that even 

 a very slight appreciation of the subject should suffice to render 

 possible some improvements of method. A first step in this direc- 

 tion was attempted by the invention of the spectroheliograph in 

 1889. The original purpose of this instrument was the photography 

 of the chromosphere and prominences, in order to simplify and ren- 

 der more accurate the daily delineation of their form. It was subse- 

 (juently found, as will be shown in the present paper, that the instru- 

 ment had a far wider range of application, and that it could be 

 applied in directions which had not suggested themselves in 1889. 



The principle of the spectroheliograph is exceeding^ simj^le. 

 Imagine a direct -vision spectroscope in which the eyepiece ordina- 

 rily employed is replaced by a (second) slit. If an image of the sun 

 is formed on the first slit of this spectroscope, the second slit will 

 permit the passage of only a narrow region of the spectrum corre- 

 sponding in width to this slit. If the slit is now moved until it coin- 

 cides with the H/5 line, for example, only hydrogen light will pass 

 through the instrument. If, then, a photographic plate is placed 

 behind and almost in contact with the second slit, and the spectro- 

 scope is moved at right angles to its optical axis, an image of the 

 sun, in monochromatic hydrogen light, will be built up on the plate 

 from the successive images of the slit. If the exposure is suitable, 

 the chromosphere and prominences will be shown surrounding this 

 image. 



Such is the spectroheliograph in its simplest form. It is obvi- 

 ously immaterial whether tlie motion be given to the spectroscope, 

 on the one hand, or to the solar image and photographic plate, on the 

 other. It is only necessary that the relative motion of the solar 

 image and first slit be such that light from all parts of the solar disk 

 shall pass successively through the slit, while the photographic plate 

 and second slit experience a corresponding relative motion. The 

 second slit serves simply to isolate any desired line in the spectrum; 



