On the Chief Line in the Spectrum of the Nebulce. 177 



In these observations, therefore, by one of our most skilled spec- 

 troscopists, we have a difference of 40 of the units now adopted, 

 and I cannot refrain from pointing out that either the diffi- 

 culties of the observations or the liability to instrumental error mnst 

 be very considerable when we see such variations as these, " a suffi- 

 ciently powerful spectroscope " and the magnificent instrument of 

 Lord Crawford's observatory being employed. 



It will be seen from this short retrospect 



(1.) That the mean of the recorded observations of the magnesium 

 fluting placed it at 5004' 7, while Dr. Huggins's last description (that 

 I had seen) of the position of the nebula line in terms of wave- 

 length gave 5008'0, as he stated it, or 6005*1, as it may be stated 

 if we take Thalen's value for the nitrogen line. These observa- 

 tions, according to his own statement, were only trustworthy within 

 a limit of sixty units, while the distances from the magnesium 

 fluting are thirty-three and four units respectively, according to 

 which measure of the nitrogen line be taken. From the facts at 

 my disposal, it was obvious that, if any difference existed, the 

 magnesium fluting was more refrangible than the nitrogen line, and 

 therefore than the nebula line, assuming the accuracy of Dr. Huggins's 

 observation of 1868. 



(2.) That, if observations by others be considered, the wave-length 

 of the magnesium fluting lies well within the extreme limits ; and, 

 indeed, not far from the mean of them all. 



From these facts, I trust it will be seen that I was perfectly 

 justified in stating the wave-length of the chief nebula line to three 

 figures only, and, further, that the coincidence between it and the 

 magnesium fluting was sufficiently probable to justify the making of 

 a statement " with reserve " to that effect. 



Since my paper of 1887, however, was presented to the Royal 

 Society, I gather from Dr. Huggins's criticisms that he has entirely 

 changed his ideas of the accuracy possible in these inquiries, and now 

 practically withdraws all the statements on which I depended to form 

 an estimate of the amount of accuracy that could be counted upon, 

 and the instrumental means that could be employed, in these 

 researches. As I have shown, the accuracy which Dr. Huggins 

 had attained with all his known skill in his last published 

 observation of the position of the nebula line was trustworthy 

 only within sixty units, according to his own statement ; this was 

 in 1868. During the last twenty years, so far as I can make out, this 

 observation has not been improved upon by the more powerful aids 

 to investigation now in his possession, while, on the other hand, as 

 recently as 1881, he regarded with complacency, as I have before 

 stated, a variation of 100 units between the measured place of the 

 same line in laboratory and observatory. Further, in all his impor- 



