58 BULLETIN 1059, U. S. DEPARTMENT OF AGRICULTURE. 



11. Finally, in following out this line of thought, the spectro- 

 bolometer represents the present limit of thoroughness. By measur- 

 ing the heat energy of every part of the spectrum, all of the results 

 may be expressed in the same absolute terms, whereas the results ob- 

 tained by photochemical reactions must be in different terms for each 

 of the several reactions which are required to cover the spectrum. 

 Furthermore, bolometric observations permit the comparison of de- 

 ficiencies in any region with deficiencies in the whole energy of the 

 light as determined by the same means. As has been pointed out, the 

 heat energy in the region of greatest chemical activity is small ; but 

 it is not too small for precise measurement, and when once measured 

 it may be transposed to terms of definite chemical reactions, the trans- 

 position factor varying, of course, with each wave-length and with 

 each reaction considered. 



The Langley bolometer has been briefly mentioned, because it was 

 designed for measurements of the whole energy of sunlight. The 

 following description from Baly (52) indicates the manner in which 

 the same principle was adapted to the most minute quantities. 



In his final work upon the solar spectrum, Langley made use of a new appa- 

 ratus 7 ; the light from a 20-inch siclerostat passed through the slit of a horizontal 

 collimator, which possessed a lens of rock salt 17 centimeters clear aperture, 

 and 10 meters focal length. This lens focused the ray upon a prism or grating ; 

 the prism was of rock salt, and was 18.5 centimeters high and 12 centimeters 

 deep in the face, and had a refracting angle of 60°. The angular width of the 

 bolometer thread was decreased to 2 inches of arc by using a telescope lens of 

 5 meters focus ; the sensitiveness was thereby increased, and by improvements 

 in the galvanometer the apparatus was made capable of detecting a temperature 

 change of 0.000001° C. The whole spectrometer was of the fixed-arm type, and 

 the spectrum was made to pass over the bolometer strip by rotating the prism. 

 An automatic self-registering method was adopted of recording the galvanometer 

 readings. The spot of light reflected from the galvanometer mirror was focused 

 upon a broad strip of photographically sensitive paper. This paper strip was 

 caused to move slowly in a vertical direction, and in this way a faithful record 

 of the excursions of the light spot was obtained. At the same time the prism 

 was slowly rotated, and therefore this record clearly showed all the temperature 

 changes of the bolometer as the spectrum passed over it. Further, the motions 

 of the sensitized paper and the prism were exactly coordinated, so that the angu- 

 lar position of the prism corresponding to any portion of the galvanometer record 

 could at once be obtained. In this way, since the dispersion of the prism was 

 already known, the wave length of any spectrum line shown upon the record • 

 could be found, and also, from the length of the throw of the light spot, its in- 

 tensity estimated. The delicacy of this apparatus was sufficient to show the D 

 lines widely separated, with the nickel line in between. * * * 



By means of this apparatus, Langley mapped the solar spectrum as far as 

 55,000 Angstrom units, and observed 700 lines between A and this limit. 



12. Evaporimeters' may be used for a very rough measure of the 

 heating value of sunlight. At first thought it would seem that the 

 rate of evaporation would be an almost ideal measure, since the 



7 Brit. Ass. Rep., 1894, p. 465; and Nature, 51, 12 (1S94). 



