July 12, 1895.] 



SCIENCE. 



35 



superficially, and there is every reason for 

 believing that the Earth, if heated to in- 

 candescence, would afford a spectrum very 

 like that of the Sun itself. 



The chemical spectra of manj^ metallic 

 elements freed from impurities are not yet 

 fully known, but these are in the process of 

 thorough investigation bj^ Eowland, and 

 Kayser and Kunge of Hanover. Their re- 

 searches will make possible a more search- 

 ing comparison with the solar spectrum, 

 hundreds of the dark lines in which are due 

 to absorption by the Earth's atmosphere, 

 and are consequentlj^ called telluric lines. 

 Especial studies of these have been made by 

 MM. Janssen, ThoUon and Cornu, Becker 

 and McClean. Whether the solar spec- 

 trum is constant in character is not known ; 

 with a view to the determination of this 

 question in the future, Professor Piazzi 

 Smyth conducted a series of observations 

 for fixing the absolute spectrum in the j'car 

 1884. Mr. Higgs, of Liverpool, studjdng 

 those strikingly marked bands in the solar 

 spectrum due to the absorption by oxygen 

 in our atmosphere, and known as ' great 

 £ ' and ' great A,' finds that the double 

 lines are in rythmic groups, in harmonious 

 sequence, capable of representation by a 

 simple geometric construction. 



Eegarding the solar spectrum (prismatic) 

 as a band of color merely, the maximum 

 intensity of heat rays falls just below the 

 red (at some distance inferior to the dai'k 

 Fraunhofer line A) ; and that of light falls 

 in the yellow (between D and E); and 

 that of chemical or photographic activity, 

 in the violet (between G and H); but in the 

 normal spectrum these three maxima are 

 brought more closely together, approaching 

 the middle of the spectrum, which nearly 

 coincides with the yellow D lines of so- 

 dium. 



Beyond the red in the solar spectrum is a 

 vast region wholly invisible to the human 

 eye ; but modern physicists have devised 



methods for mapping it with certainty. 

 Sir John Herschel, J. W. Draper and Bec- 

 querel were the pioneers in this research, 

 the last utilizing various phosphorescent 

 substances upon which an intense spectrum 

 had been projected for a long time. Direct 

 photographic maps of the infra-red regions 

 are very difficult, because the actinic in- 

 tensity is exceediag feeble ; and Abnej^, by 

 means of collodion plates specially prepared 

 with bromide of silver, has made an ex- 

 tended catalogue of the invisible dark 

 bands. But Professor Langley has pushed 

 the mapping of the infra-red spectrum to 

 an unexpected limit by means of the bolom- 

 eter, a marvellously sensitive energy-mea- 

 surer of his own invention. In order to un- 

 derstand in outline the operation of the 

 bolometer, or spectro-bolometer, it is nec- 

 essary to recall that, as the temperature 

 of a metal rises, it becomes a poorer con- 

 ductor of electricity; as it falls its conduc- 

 tivity increases, iron at 300° below centi- 

 grade zero being, as Professor Dewar has 

 shown, nearly as perfect an electrical con- 

 ductor as copper. The characteristic fea- 

 ture of the bolometer is a minute strip of 

 platinum leaf, looking much like an exceed- 

 ingly fine hair or coarse spider web. It is 

 about ^ inch long, yjjj inch broad, and so 

 thin that a pile of 2.5,000 strips would be 

 only an inch high. This bolometer film, then, 

 having been connected into a galvanometer 

 circuit, is placed in the solar spectrum 

 formed either by a grating or through the 

 agency of rock salt prisms ; and as it is 

 carried along the region of the infra-red, 

 parallel to the Fraunhofer lines, the fluctu- 

 ations of the needle maj^ be accurately re- 

 corded. 



In this manner he first represented the 

 Sun's invisible heat spectrum in an energy- 

 curve; but his recent application of an in- 

 genious automatic method, accessory to the 

 bolometer, has enabled him to photograph 

 its indications in a form precise^ compar- 



