of the Solar Specti-um. 247 



the undular length 00024, falls beyond the boundary S of the 

 solar spectrunij which corresponds to the least refrangibility. 

 Altogether, therefore, the solar spectrum contains rather more than 

 2^ octaves, that is, rays whose undular length is betiveen 0*0003 

 and 0'0018 millimetre. 



Distribution of heat in the diffraction-spectrum. 



It is well known that the distribution of the colours in a 

 " grid"-spectrum are quite differently arranged from those of a 

 prismatic spectrum. It may therefore be predicted that the 

 thermal curve of intensity for a grid-spectrum would have quite 

 a different form from that given in fig. 1, which is the curve 

 for a prismatic spectrum. 



As far as I am aware. Draper was the first who examined 

 the heat in a diffraction-spectrum. He experimented upon a 

 gi"id-spectrum obtained by reflexion*. The manner in which 

 he conducted his experiments is exceedingly imperfectly de- 

 scribed; nor does he give the slightest account of the maguitude 

 of the effects which he obtained by the multiplier. He only 

 asserts that he found the maximum heat in the yellow. Al- 

 though, as he himself admits, his results ai'e "imperfect and 

 incomplete," still a more exact description of his apparatus, the 

 ari'angement of his experiments and conclusions, would have 

 been of interest to physicists. 



If I undei'stand rightly. Draper employs as thermo-battery a 

 single thermo-electrical element. If this be so, it is no wonder 

 that the thermal effects which he observed were, as he himself 

 admits, exceedingly small. Indeed, according to my own expe- 

 rience, I can scarcely comprehend bow he could observe more 

 than traces of such effects. 



Being in possession of an excellent " grid," similar to that 

 with which Eisenlohr experimented f, and for which I am in- 

 debted to the kindness of Professor Schwerd, I commenced ex- 

 periments on the radiant heat in the diffraction-spectrum. I 

 found, however, although employing a thermo-battery of fifteen 

 pairs, that the thermal effects were so small that I relinquished 

 the hope of obtaining useful results in this manner; for ap- 

 preciable deflections of the multiplier could only be obtained by 

 bringing the thermo-battery so close to the grid that the two 

 spectra were so impure that they were completely confounded. 



The small amount of heat in the diflVaction-spectrum is 

 easily explained. If no lens be employed, not more than 

 2 miilims. broad of the grid can be used, if anything like pui-e 



* Phil. Mag. 1857, vol. xiii. p. 153. 

 t Pogg. Ann. vol. xcviii. p. 354. 



