KEPORT OF THE SECRETARY. 79 



A display of transparencies showing apparatus, buildings, and results of observa- 

 tion is a prominent object to visitors as they enter the room, and thence they pass 

 through the darkened chamber above mentioned and get a view of the solar spec- 

 trum and of such sun spots as are Adsible. As they leave the darkened chamber they 

 come upon a group of bolometric apparatus in actual operation and sensitive enough 

 to give 100 or more scale divisions deflection when the visitor holds his hand in 

 front of the bolometer. Charts and reports illustrative of the work of the Observa- 

 tory are found upon tlie walls. A large number of copies of a pamphlet descriptive 

 of the Observatory and its studies have been freely distributed. 



2. Improvements of apjmratus. 



Bolometer. — The bolometer and its adjuncts had reached so high a state of perfec- 

 tion, as described in my last year's report, that further improvement could hardly be 

 expected, but in order to make it practicable to set up so sensitive an instrument at 

 the St. Louis Exposition and leave it without expert attention for months, it was felt 

 desirable to coml)ine in comi)act form the most approved methods of construction, 

 having special reference to simplicity of manipulation, so that the instrument could 

 be safely left in unskilled hands. Accordingly the bolometric apparatus, shown in 

 Plate VI, was designed at the Observatory and constructed in the Observatory shop. 



As in the form shown in Plate XII A of Volume I of the Annals of the Astro- 

 physical Observatory, all the adjuncts to the bolometer, excepting the battery and 

 galvanometer, are combined in a single case, adjusted mechanically from outside, but 

 in this new form tlie adjusting slide wires are straight, instead of spiral, and no 

 clamping mechanism is required, so that the construction is nmch simplified and can 

 hardly get out of order. At the same time five slide wires are provided of differing 

 sensitiveness of adjustment, so that by merely turning cranks on the outside of the 

 case any adjustment from 2^0 to 2 00 off 0000 i^^ the relative resistances of the l)alancing 

 arms of the bolometric circuit can be easily effected. It is a principal advantage of 

 this arrangement that all the electrical circuit is in such compact quarters that 

 changes of temperature affect all parts almost alike. These have been the great causes 

 of disturbance of the deflections of the needle in the past, and the great obstacle to 

 an automatic registry. Their bad effects are so nearly eliminated now that apart 

 from the occasional need of a uniformity in the temperature of the prism, the 

 elaboi-ate arrangements for keeping a uniform temperature in the observing room are 

 no longer needed. No difficulty from drift or any disorder has ever been experienced 

 with this adjunct to the bolometric apparatus, and it is so much superior to any 

 earlier form that a duplicate piece has since been constructed for use in a research 

 on the radiation of the stars. Both were made at the Observatory shop and worked 

 perfectly from the very first trial, and they reflect great credit on Mr. Kramer, the 

 instrument maker. 



Pyrheliometer. — Reference was made in last year's report to the new form of absolute 

 pyrheliometer then being developed, consisting of a hollow chamber or "absolutely 

 black body" in which the radiation is absorbed and from which the heat is continu- 

 ously removed by a liquid circulating about' the walls of the chamber. This instru- 

 ment has been tried so successfully in an exjierimental form that steps are now being 

 taken to provide one for continuous automatic registration of the rate of solar radia- 

 tion. Its principal advantage as an absolute instrument depends on the fact that it 

 may be demonstrated that its indications are correct, for if a known quantity of heat 

 is supplied electrically to a coil within the chamber, this heat will reach the walls by 

 convection and radiation, and being then removed by the flowing liquid its amount 

 may be measured and compared with the known heating actually produced in the 

 coil. But it is obvious that the heat of solar rays, absorbed almost wholly upon the 

 wall of the chamber as they fall upon it, is much more likely to l)e accurately meas- 

 ured than the heat of the coil, which must first be chiefly communicated to the air 



