THE NEW SPECTRUM. gg9 



how far this spectnun extended ncr. perhaps, what these explorations 

 really meant. They may be compared to actual journeys into this 

 dark continent, it it may be so called, which extended so' far beyond 

 those of previous explorers that the determination of position, bv the 

 writer, corresponding- somewhat to longitudes determined by the terres 

 trial explorer in a new country, was, by those who had not bee.i so far 

 but had conceiyed an inadequate idea of the extent of the recrion 

 treated as erroneous and impossible. " 



A necessary limit to the farthest infra-red was in 1880 supposed toexist 

 near the waye-length l^ Doctor John Draper,^ for instanc,^ announced 

 in other terms that the extreme end of the inyisible spectrum might, 

 from theoretical considerations, be probably estimated at something 

 less than the waye-length of l^ whence it followed that the above 

 value of 1.8'^ was impossible, and, still more, that of 2.8*^. If, in this 

 connection, we revert to our map, where the visible spectruni has an 

 extent in wave-lengths of 0.36'^, then, on that same scale, the length of 

 the entire possible spectrum, visible and invisible^ was fixed by Draper 

 at the point there shown near the band pffr. In still other words, 

 according to him the very end of any spectrum at all would be less 

 than 3 on a scale in which the visible spectrum was 1. Doctor Draper's 

 authority was deservedly respected, and this citation of his remarks is 

 made only to show the view then entertained by eminent men of 



science. 



Now, the writer had proved by actual measui'ement that it extended 

 far beyond this point, and had announced, as the result of experiment, 

 that it extended at any rate to about three times the utmost length then 

 assigned from theoretical reasons })y Draper, founded on the tiieu uni 

 versally accepted formula of Cauchy, which was later discredited l)y 

 the direct experimental evidence giv^en of its falsit}^ by the bolometer. 



The bolometer, which is wholly Independent of light as a sensation 

 and notes it only as a manifestation of energy, fii-st lays down the 

 spectrum by curves of energy from which the linear sp(>cti-um is in 

 turn derived. Two such curves taken at different times are given to 

 show the agreement. There must now be explained, however biietly. 

 the way in which these energy curves, which are the basis of all. lia\e 

 actually been produced here. 



In making the map of the energy curves it should l)c rcnienibcrcd 

 that when an invisible band or line is suspected, its presence is revealed 

 by the change of temperature in the bolometer strijxs affecting the 

 needle of the galvanometer, causing this needle to swing this way or 

 that; let us suppose to the left if from cold and the right if from heat. 

 The writer's first method was to have one person to note the exposure, 

 another to note the extent of the deflection, and a third to note the 



1 Proceedings of the American Academy, vol. xvi, p. 233, 



1880. 



