Appendix V. 



EEPOET ON THE WORK OF THE ASTROPHYSICAL OBSERVATORY FOR 

 THE YEAR ENDING JUNE 30, 1896. 



Sir : The bolometric investigation of the infra-red solar spectrum has, under your 

 general direction, continued to he the main feature of the work of the Observatory 

 for the past year. While it has heen found necessary to postpone a final account of 

 the result of this long investigation, yet it is confidently believed that such a pub- 

 lication can be made before many months more; and that the limit of accuracy 

 in determining the position of absorption lines, which was mentioned in the report 

 for the last year as the aim of endeavor, will be reached in the results shortly to be 

 communicated. 



The year's prosecution of the research will, for convenience, be discussed under 

 the following heads : 



A. Measurement and elimination of "the drift" 1 and other sources of error 



in the spectrobolo-metric processes. 



B. General spectrobolographic work. 



C. Accessions of apparatus. 



A. Measurement and Elimination of Sources of Error in the Spectrobolo- 

 metric Processes. 



As detailed in the report for the year ending June 30, 1895, it was then evident 

 that the holographic curves were so encumbered with minute deflections due to earth 

 vibrations and extraneous electrical and magnetic disturbances, that the small and, 

 it was presumed, very numerous deflections corresponding to fine absorption lines 

 were greatly obscured. One of the means proposed for eliminating false and 

 retaining true lines has been the process of composite photography. The method 

 of application of this process had been as follows : 1. All the deflections in a series 

 of curves of a greater magnitude than a certain arbitrary amount were by the 

 "cylindric process" drawn out into lines. 2. The resulting "cylindrics" were suc- 

 cesstvely exposed before a single plate to give a composite. 



It must be remarked the process of experiment has shown that this procedure is 

 not free from objection. While it is unquestionable that deflections great enough to 

 be surely recognized as genuine will be brought out, owing to the greater prominence 

 given them in "blocking in" the curves preparatory to making the " cylindrics, " 

 yet many of the genuine deflections of the same size as the accidental ones (for dis- 

 tinguishing which the process is chiefly useful) may possibly be eliminated; for 

 owing to the various sources of error in the bolometric processes, particularly the 

 superposition of accidental deflectious over the true ones, to the personal error in 

 blocking in, and to the errors introduced in the photographic processes, slight lateral 

 displacements of the true lines in the separate cylindrics will always occur. Thus 

 it may frequently happen that in the final composite the less prominent lines, being 

 thus variable in their position to an amount equal to their own width, or even more, 

 may be eliminated. Lines corresponding to accidental deflections may, on the other 

 hand, be retained, since their number is very great in comparison with the number 

 of cylindrics employed in making composites, for it may thus frequently happen 



J It will be remembered that "the drift" has hitherto been a principal source of 

 trouble in these researches. It consists in a slow progressive movement of the 

 needle due to a great variety of extraneous causes. 

 68 



