KEPORT OF THE SECRETARY. 



77 



I am pleased to state that by the use of the improved apparatus, to which I refer 

 later, the amount of the probable error has been greatly reduced, and the investiga- 

 tion of the causes of false deflections has justified the belief that those not directly 

 due to the site of the observatory will be, ere long, eliminated. 



During thf> progress of this work the making of observations with the apparatus 

 on hand has not been suspended, but, on the contrary, has produced a larger num- 

 ber of holographs than during any preceding year. The work in detail may most 

 conveniently be classified under three general heads: 



A. General spectro-bolographic work. 



B. Special spectro-bolographic work. 



C. Improvements in apparatus and methods. 



A. The subjoined table shows in detail the number of working days and the num- 

 ber of bolographs made in the observatory during the year. 



Date. 



Days 

 available. 



Bolo- 

 graphs 



of 

 spectra. 



Experi- 

 mental 

 bolo- 

 graphs. 



Remarks. 



1894. 



4 



5 





Improving apparatus. 



Main observatory closed twenty-one days. 









5 



9 



4 



12 



7 

 8 

 3 

 8 

 12 

 5 



12 



19 



9 



33 



20 

 24 

 10 

 42 

 44 

 14 







1 











2 

 1 





1895. 













Constructing new piers. 





1 



20 

 4 









Unusually bad weather. 







77 



232 



29 



29 





Total 



261 











On the 5th of December, 1894, an extended experimental investigation of the 

 errors exhibited by the apparatus and the methods in use was begun on a scale not 

 before tried, and the effects from variations in the potentials of the batteries and 

 from temperature effects in the complicated circuit of the bolometer, as well as 

 from the tremors of the earth and changes of the magnetic field, incident to the bad 

 location of the observatory, were found to be larger than had been assumed from 

 previous trial. Since that time the reduction of these errors has been the constant 

 effort of the observatory. 



The entire circuit of the bolometer has been overhauled and made to consist, as 

 nearly as possible, of copper of the same density and without joints which might 

 produce thermal effects. 



The experimental bolographs, referred to in the above table, were, for the most 

 part devoted to the investigation of the error produced by minute variations in the 

 potentials of the batteries used in the bolometer circuit. In order to give some idea 

 of the delicacy of the performance required of this apparatus, it may be added that 

 the variations averaged only 0.0000007 of a volt, but were still considered to be too 

 large for accurate working. It has also become evident that the arrangement of 

 balancing coils, which have up to the present time been of the usual laboratory 

 type, were capable of producing, under the fluctuations of temperature in the 

 observatory, effects of the magnitude of those produced by the minute solar lines, 

 which it was now being attempted to define and measure. 



