170 Mr. J. N. Lockyer. 



a scale in the measurements made in the laboratory and observatory 

 respectively of a line which his diagram shows as clearly defined, 

 though in the text he uses the ambiguous phrase that he was able to 

 see a " distinct impression " of it. He did not consider that a 

 discrepancy of the magnitude indicated threw any doubt upon the 

 identity of the lines. 



I may mention as another fact which supports the use of a three- 

 figure reference, that even in the research on the brightest stars it 

 was difficult to absolutely reconcile the one prism work in the 

 observatory with laboratory work. No better illustration of this 

 could be found than a comparison of the work of two such observers 

 as Dr. Huggins and Prof. Cornu. 



One of the chief points in the memoir was the discovery of a series 

 of lines in the ultra-violet which Dr. Huggins ascribed to hydrogen ; 

 lines near the position given have, indeed, since been measured in 

 hydrogen by Cornu.* I append the wave-lengths as given by the 

 two observers : 



Huggins.t Cornu. 



3767-5 3769-4 



3745-5 .... 3749-8 



8730-0 .... 3733-6 



3717-5 .... 3720-6 



3707-5 .... 37iO-7 

 3699-0 



It will again be observed, that if one had wished to give a handy 

 reference to these lines a short title three figures would have been 

 sufficient in two cases, for we have 



3717-5 .... 3720-6 

 3707-5 3710-7 



It is no part of my present business, however, to discuss the relative 

 accuracy of Dr. Huggins and Prof. Cornu as observers, but it must be 

 pointed .out that by the nature of the research Prof. Cornu is more 

 likely to be right. We owe one of our best maps of the solar spectrum 

 in this part to Prof. Cornu, and the comparison of hydrogen with the 

 Sun could always be repeated at leisure and under stable conditions, 

 whereas, in the case of Dr. Hugging, the result depended upon a 

 photographic solar comparison in the telescope taken some hours 

 afterwards. In any case, Professor Cornu's numbers are three years 

 old, and, so far as I know, Dr. Huggins has not challenged them. 

 An inspection of the numbers shows that there is, in all probability, 



* ' Journal de Physique,' vol. 10, 1866, 341. 

 t ' Phil. Trans.,' vol. 171, p. 682. 



