REPORT OF THE SECRETARY. 73 



constant source of heat has been found, and all the variations of the source employed 

 are included in this 0.035 of 1 per cent. In the earlier use of the bolometer such 

 precision would have been unattainable, owing to the nonstability of the zero point 

 of the galvanometer, but the "drift" and tremor have now been so far eliminated 

 as to make such results quite possible. 



ACCESSIONS OF APPARATUS. 



There have been added to the equipment of the Observatory during the year the 

 following considerable pieces of apparatus : 



One Richard barograph. 



One Crova actinometer. 



One Fuess spectrometer, with 17 centimeter circle reading to 10 seconds. Acces- 

 sories consisting of reading and collimating telescopes, micrometer slit, and one 

 liquid prism. 



One reflecting galvanometer of the Thomson type, already referred to. Case con- 

 structed at Astrophysical Observatory shop; coils and needle at Astrophysical 

 Observatory. 



Two cooling tanks and ammonia compression apparatus, by De LaVergne Refrig- 

 erating Company. 



Automatic temperature control for the above, by Johnson Temperature Regulating 

 Company. 



One " hot box" provided with equatorial mounting. 



Two cylindric collimating minors by Brashear, already referred to. 



Ten cells "Cupron" battery, Type I, from Umbreit & Mattb.es, of Leipzig, for use 

 on the bolometer circuit. 



A device for floating galvanometer upon mercury, constructed at the Astrophysical 

 Observatory shop. 



One ball-and-socket mounting for salt cylindric lens to close the bolometer case 

 air-tight, while allowing of the adjustment of the lens for best definition. Con- 

 structed at the Astrophysical Observatory shop. 



Two salt cylindric lenses for the above mounting, by Kahler. 



One adjustable mounting for slit of spectroscope. Constructed at the Astro- 

 physical Observatory shop. 



One air tight galvanometer case. Designed at the Astrophysical Observatory. 

 Constructed by Gaertner & Co. 



One bolometer case, by Gaertner & Co. 



Besides these pieces of apparatus actually received, there were ordered the fol- 

 lowing: 



Fourteen cells of •' Cupron " battery, Types I and III, with extra parts. 



One Rubens thermopile. 



One combined bolometer and rheostat after new design already referred to. 



PERSONNEL. 



The services of Mr. C. E. Mendenhall, as assistant, were secured for the period of 

 three months beginning June 1, 1898. 



SUMMARY. 



In conclusion I may say that the investigations of the absorption bands in the 

 infra-red solar spectrum, reopened by the securing of more highly sensitive holo- 

 graphic apparatus during the unavoidable delay in publication of the results 

 attained last year, has been attended this year with a degree of success exceeding 

 anything Avhich could be hoped for. The result now reached, which will undoubtedly 

 be ready for the press early in the coming calendar year, includes the discovery and 



