Oxygen in the Sun. 309 



element which is reproduced in the corresponding lines in the 

 solar spectrum. Thus the nebulous character of the lines of 

 magnesium is perfectly reproduced in the corresponding lines 

 in the solar spectrum. The test of coincidence, therefore, re- 

 quires primarily a normal spectrum, and the highest possible 

 dispersion. The earlier observers were limited to instruments 

 of small dispersion, and the entire number of lines observed 

 in the solar spectrum was small compared with that given by 

 the best modern apparatus. The chances for an apparent 

 coincidence were therefore much greater, and evidence of a 

 very misleading character could be obtained. 



In Dr. H. Draper's published photograph, the coincidence 

 of the greater part of the oxygen lines with bright bands in 

 the solar spectrum is quite striking ; and it is not a matter for 

 surprise that he was led to conclude the connexion between the 

 two spectra to be a physical one, and to announce the existence 

 of oxygen in the sun as proved. Instances are not infrequent 

 where instrumental imperfection or lack of power has led to 

 results unsupported by later and more powerful research. 

 Witness the spots of Venus of the older observers. Now 

 when we apply to the spectra of the sun and oxygen a dis- 

 persion and definition that show the minute detail of each, the 

 " bright bands v at once vanish, or no longer appear as such, 

 and all the apparent connexion between them and the oxygen 

 lines disappears also. The bright bands of Dr. H. Draper's 

 spectrum are found to be occupied by numerous dark lines, 

 of various degrees of intensity ; but the hypothesis of Prof. 

 J. C. Draper, that these are the true representatives of the 

 oxygen lines, is rendered untenable by the lack of any syste- 

 matic connexion between the two. It happens quite fre- 

 quently that an oxygen line falls centrally upon the space 

 between two dark lines of the solar spectrum, but not more 

 frequently than we might expect as a matter of chance, when 

 we consider the vast number of lines and spaces ; and the 

 fact that the spaces are no brighter than the surrounding 

 background of the solar spectrum would not seem to permit 

 of their interpretation as bright lines. 



The subject of bright lines in the solar spectrum is one 

 upon which men will probably differ, and we have sought 

 information upon it. Of course there is no a priori reason 

 why such bright lines should not exist, as they do in many 

 stars ; but we have photographed the sun's spectrum every 

 day that the sun has shone for nearly five months, without 

 finding a line that could with certainty be pronounced 

 brighter than its neighbours ; and it must be admitted that 

 the photograph is the best of photometers in such a case. 



