KEPORT OF THE SECRETARY. 79 



a more perfect automatic regulation of the temperature (in the interest of the work 

 rather than of the comfort of the observer) is being provided for. 



During the year 1893 the galvanometer, perhaps the most important part of the train 

 of apparatus, was brought to such a condition that no further improvement in this 

 direction was regarded as necessary until justified by a more permanent building 

 and better facilities for isolation from magnetic and seismic disturbances than are 

 at present capable of being obtained, but in order to make the best use of the facili- 

 ties at hand, the former unstable galvanometer pier has been replaced by a massive 

 brick and stone structure, finished March 23, 1895, built in such a manner that pro- 

 tection from the vibrations of the surface earth is insured. It is, however, proper 

 to repeat that the vibrations of the earth, as well as the seismic disturbances, conse- 

 quent to the location of the observatory amid the traffic of the city, continue to pre- 

 sent serious causes of error which no possible care or pains seems able to prevent. 



By examinations conducted in October, 1894, it was shown that the driving clock 

 of the spectro-bolometer train performed its part with less than the rigorous accuracy 

 required. It has consequently been replaced by a much more accurate mechanism, 

 of which an illustration is given, installed April 2, 1895, which was constructed by 

 Messrs. Warner & Swasey, of Cleveland, in part, and in part by Mr. John A. Brashear, 

 of Allegheny. This clock which controls both the movement of the photographic 

 plate and of the great azimuth circle carrying the prism, has been shown, under the 

 strain of repeated use, to perform its duty within the limit of allowed probable 

 error — that is, one-tenth of a millimeter in 60 centimeters motion of the plate, or six- 

 tenths of a second of arc in one degree at the circle. A new pier has also been 

 constructed for the support of this clock. 



A mean solar standard clock, by the Howard Watch and Clock Company, has been 

 placed in the basement of the Smithsonian Institution. 



Great difficulty has always been experienced in the proper protection and preser- 

 vation of the optical surfaces of the rock salt necessarily employed. A large gain 

 in time and economy has been made by introducing the practice of repolishing the 

 surfaces in the observatory itself, and it is now possible to refigure the surface of a 

 large prism or lens in a few hours, at a small cost, where it was formerly the custom 

 to send such a prism or lens to distant opticians at a considerable expense of time 

 and money. 



A new bolometer case, with provision for receiving radiations from two sources, 

 and many minor pieces of apparatus, have been constructed in the workshop of the 

 observatory. These improvements in the apparatus have been accompanied by 

 improvements in the method of their use, which will more fittingly form the subject 

 of future report, but they may nearly all be connected with the need of a more con- 

 stant temperature, combined with freedom from ground and magnetic tremors. 



MINOR WORK OF THE YEAR. 



Some further experimental data were collected for the more accurate determina- 

 tion of the laws governing the change of the index of refraction of rock salt, with 

 changes in its temperature. The approach of observing weather prevented the com- 

 pletion of the experiments. 



All of the bolometers on hand have been calibrated, in order that their relative 

 efficiencies, as well as the means of reading their indications quantitatively, might 

 be known. 



The very minute and rapid changes in the potential of certain types of storage 

 batteries, intended for the use with the bolometer, have been measured, and their 

 effect upon the indications of the bolometer investigated. 



It should finally be repeated, as the most important conclusion of the year, that 

 the final degree of precision attainable can never be reached in the present site. 



PERSONNEL. 



Mr. C. G. Abbot was appointed as assistant in the observatory on the 19th of June, 

 1895. 



