S. P. Langley on the Selective Absorption of Solar Energy. 155 



evidence the subject admits of; that every ray, whether lying 

 in the "chemical/' "visible," or "heat" region, is capable of 

 making itself known as heat, and that the maximum of heat 

 in the normal spectrum is near the yellow. 



Further, by taking all the observations twice daily, at times 

 when the atmospheric absorption is very different, we are able 

 to calculate (for the first time) the amount of this absorption 

 for each separate spectral ray. These researches are neces- 

 sarily long and difficult ; but they have led to some very 

 unexpected results. The reader who wishes to pass at once to 

 these results will find them in the summary at the close of 

 this memoir. 



Preliminary Observations. 



The measurements with the grating possess the inestimable 

 advantage of enabling us to fix the wave-length of every ray 

 measured; but while the heat in the grating-spectrum is, as 

 has just been said, at best less than one tenth that in the pris- 

 matic, the latter is itself, when taken in portions so narrow as 

 to be approximately homogeneous, almost insensible. The 

 difficulties of measuring heat with the grating are greatly 

 complicated by the overlapping of the spectra. In these 

 first measures, which were carried to a wave-length of one 

 thousandth of a millimetre*, I have employed two of the 

 admirable gratings of Mr. Rutherfurd — one containing 17,296 

 lines to the inch, or 681 to the millimetre, and one half that 

 number, both ruled upon speculum-metal. 



I have used a slit at a distance of 5 metres, without any 

 collimator, and have kept the grating normal to the optical 

 axis. It will be seen, then, that the rays have passed through 

 no absorbing medium whatsoever, except the sun's atmosphere 

 and our own. 



The rays from the grating fall upon a concave speculum 

 (whose principal focal distance is about 1 metre), and from 

 this are concentrated upon the mouth of the bolometer, form- 

 ing a narrow spectrum, which passes down the case of the 

 instrument and falls upon the bolometer-thread. As this 

 thread moves along the spectrum parallel to the Fraunhofer 

 lines, its coincidence with one of them is notified by a lowering 

 of its temperature and a deflection of the galvanometer. The 

 instrument is of course equally sensitive to the invisible radia- 

 tion as to the visible. It is important to observe that no 



* Throughout these measures the unit of wave-length will be the micron 

 if 1 ) — ToVo millimetre, or 10,000 times the unit of Angstrom. Thus the 

 wave-length of Fraunhofer's U A" is here written M, 76. 



M2 



