466 REPORT— 1894. 



first investigator here was Sir William Herschel, whose observations con- 

 sisted essentially in finding that there was something which the eye could 

 not see in a region which he proposed to call the ' thermometric spectrum.' 

 His distinguished son, Sir John, made a curious anticipation of later 

 discovery, by indicating, though crudely, that this invisible heat was 

 not uniformly distributed, and a similar conclusion was reached in an 

 entirely different manner, through the thermopile, by the too-early-l<ist 

 Melloni. So ignorant, in spite of these investigations, of those of the 

 elder Draper, and of the elder Becquerel, were we till lately, that when, 

 quite within my own recollection and that of most of you, Lamansky in 

 1871 published from his obsei'\'ations with the thermopile a crude little 

 illustration showing three inequalities in the energy curve, universal 

 attention was excited by it among those interested in the subject. 



Among other minds, my own then received a stimulus which turned 

 in this direction, and having, as it seemed to me, exhausted the capacities 

 of tlie thermopile, I invented an instrument for continuing the research, 

 which was afterwards called the bolometer, and with which in 1881, at an 

 altitude of 13,000 feet upon Mount Whitney, I found spectral regions 

 hitherto unreached, and whose existence had not been suspected. 



I returned with a strong impression of the prospective importance of 

 this discovery, and laboured at the Alleghany Observatory in improving 

 all portions of the new method of research, especially of the bolometei' and 

 its adjuncts, with the twofold object of obtaining greater sensiti\-eness 

 to heat, and greater precision in fixing the exact point in the spectrum 

 where the change of heat originated. With the former object such a 

 degree of sensitiveness was at that time reached, that the bolometer 

 indicated a change of temperature of yoo otto of '>' degree Centigrade, and 

 with the latter, such precision, that it was possible to fix the relative 

 position of a line not merely with a possible error of a considerable 

 fraction of a degree, such as Lamansky's determination is evidently subject 

 to, but with a certainty that the error would be within a minute of arc. 

 The range of the apparatus in wave-lengths was almost unlimited as com- 

 pared with any other process, and both its sensitiveness and its possible pre- 

 cision seemed to be at that time notable as compared with previous methods, 

 for a great advance was made on anything done before with the thermo- 

 pile, when the presence of the well-known 'D' line of sodium was rendered 

 sensible (though barely sensible) even as a single line by the change of tem- 

 perature. This sensitiveness was also, as has been said, accompanied with 

 the possibility of unusual precision. The results of this labour were laid 

 before the British Association in the communication already alluded to, 

 and which exhibited ten or twelve inflections of the curve in the portion till 

 then almost unknown, which extends from a wave-length of 1" to a 

 wave-length of nearly 3", at which point the glass prism then used became 

 wholly opaque to radiation. The positions of these inflections were fixed 

 with a precision quite impossible to the thermopile, but this exactness was 

 only obtained in practice by a process so slow as to be almost prohibitory ; 

 and with this apparatus I made in those earlier years such a number 

 of observations as I hardly like to recall, so disproportionate did the 

 labour inherent in this method seem to the final result. 



The justification of this labour seemed to lie in the fact that it does not 

 appear that photography has ever rendered anything much below a wa\e- 

 length of I'' — anything at least which has lieen reproduced for publica- 

 tion in a way which gives confidence that we are in touch with the original. 



