OF ARTS AND SCIENCES. 



297 



Conclusions. 



It will be observed that the corrections to be applied to the wave- 

 lengths obtained by Liveing and Dewar are progressive in their nature 

 when compared with those which must be applied to the results of 

 Hartley and Adeney. The difficulty in identifying lines and deter- 

 mining coincidences by employing the tables of metallic spectra in the 

 ultra violet, published by the British Association, is illustrated by our 

 work ; for certain lines measured by Liveing and Dewar, which are 

 identified by the committee with lines given by Hartley and Adeney, 

 are in reality removed from each other, one or two lines intervening. 

 In certain cases the lines of Liveing and Dewar are wholly beyond 

 identification with those given by Hartley and Adeney. 



The results of our inquiry into the accuracy of the results of previ- 



