February 16, 1906.] 



SCIENCE. 



263 



seum work. The chapter on ' The Mineralog- 

 ical and Geological Cabinets,' by D. S. Mar- 

 tin, contains many interesting references to 

 Cooper, Le Conte, Holmes and others of our 

 earlier mineralogists and geologists. 



SOCIETIES AND ACADEMIES. 



THE PHILOSOPHICAL SOCIETY OF WASHINGTON. 



The 610th meeting, the 35th annual meet- 

 ing, was held on December 30, 1905, President 

 Littlehales in the chair. 



The report of the secretaries showed 126 

 members on the active list, a small net gain 

 during the year, and 74 on the absent list. 



The report of the treasurer showed an in- 

 come of $802.19 and expenses of $413.27 dur- 

 ing 1905. 



The following officers were elected for the 

 year 1906: 



President — Cleveland Abbe. 



Vice-Presidents — ^A. L. Day, E. B. Rosa, L. A. 

 Bauer and J. F. Hayford. 



Treasurer — Bernard R. Green. 



Secretaries — C. K. Wead and G. K. Burgess. 



General Committee — W. A. DeCaivelry, W. S. 

 Eiehelberger, L. A. Fischer, C. Adler, R. A. 

 Harris, J. Page, C. G. Abbot, L. J. Briggs and 

 I. Winston. 



To this committee there will be added such 

 of the past-presidents residing in Washington 

 as shall consent to serve on it. 



The 611th meeting was held on January 13, 

 1906, President Abbe in the chair. 



Mr. C. G. Abbott read a paper on ' A Stand- 

 ard Pyrheliometer and its Use on Mt. Wilson 

 in California.' 



Three independent lines of research are be- 

 ing carried on by the Smithsonian Astrophys- 

 ical Observatory to determine if the output of 

 solar radiation is variable. One of these con- 

 sists in obtaining values of the ' solar con- 

 stant ' of radiation outside our atmosphere, 

 by the spectro-bolometric method of homo- 

 geneous rays practised many years ago by 

 Mr. Langley on Mt. Whitney. The author 

 was in charge of a Smithsonian Observatory 

 expedition to Mt. Wilson in California to 

 determine solar constant values during the 

 past summer and autumn, and had with him a 



complete duplicate of the outfit simultaneous- 

 ly used for the purpose in Washington. 



Measurements by the pyrheliometer or 

 actinometer of the total radiation reaching 

 the surface of the earth are necessary, as well 

 as spectro-bolometric observations. Distrust 

 was entertained of the accuracy of all instru- 

 ments heretofore proposed as standard actin- 

 ometers or pyrheliometers, for the reason that 

 the rays are always received upon a front, or 

 outside surface, while the measurements of 

 temperature are made behind or within. 

 Hence the absorbed heat has a path of direct 

 escape to the surroundings and no allowance 

 for this can be made either by cooling cor- 

 rections, by reading with the temperature re- 

 cording apparatus at the temperature of the 

 surroundings or otherwise. In illustration, 

 the nature of the error in the instruments of 

 Pouillet, Angstrom, Nichols and Hull and 

 others, was pointed out. 



No easy method of determining the magni- 

 tude of the error being found, a new instru- 

 ment in which the rays are absorbed at the 

 conical rear end of a tube-like blackened and 

 diaphragmed chamber was devised. This 

 chamber is approximately the ' black body,' 

 or perfect absorber, of Kirchhoff, so that no 

 correction for reflection is needed. On ac- 

 count of its shape there is great hindrance to 

 the escape of heat by radiation or convection, 

 so that the heat will almost wholly be retained 

 somewhere on the walls. The chamber walls 

 are bathed by a spiral current of water whose 

 difference of temperature before and after 

 passage is determined by a platinum ther- 

 mometer. To assure that no heat is lost, a 

 known current of electricity can flow through 

 a coil of wire near the rear within the absorb- 

 ing chamber, and this known heating can be 

 determined as if it came from the sun. The 

 mean of nine comparisons made in this way 

 on Mt. Wilson indicated 100.4 per cent, of the 

 heat introduced found, with a probable error 

 of less than half per cent. Sun heat is more 

 favorably received than coil heat and should 

 be more exactly measured. 



The instrument is naturally perfectly con- 

 tinuous in its action, and was mounted equa- 

 torially on Mt. Wilson, the galvanometer 



