of Heat in the Prismatic Spectrum. 4.51 



prism, I wetted the dull side with spirits-of-wine. 1 thereby 

 obtained a brighter spectrum; but at the same time the red, re- 

 ceived at a distance of 6 feet, immediately fell lower by 2 Paris 

 lines than it had done before. The same circumstance, then, 

 must have occurred with the bright prism; perhaps the limit 

 of the red expanded in it still lower, and its centre might have 

 fallen just on that point where with the ground prism we 

 found the maximum of heat, but only a faint reddish glimmer- 

 ing, which is usually not reckoned when we fix the limit of the 

 coloured spectrum. But I could not arrive at certainty on this 

 point by exact measurement, for want of a proper apparatus. 



Fi'om these experiments 1 believe 1 ma}^ conclude that the 

 deepest red of the spectrum was weakened to such a degree by 

 the ground surface that it no longer formed a fixed limit, and 

 the spectrum necessarily appeared shorter than it does with a 

 polished prism. 



What, then, was effected here by external circumstances, 

 must with the flint-glass prism have been produced by inter- 

 nal; for its polish'was very good, and the colours in the spec- 

 trum very brilliant and well-defined. Nor was it deficient in 

 that weak reddish glimmering beyond the limit of the brilliant 

 red which is produced by every prism ; indeed it seemed to me 

 to be rather brighter with the flint-glass prism than with any 

 other I have used. Of this glimmering, however, I shall say 

 more below. — Whether the transfer of the maximum of heat 

 with the flint-glass prism, into this weakly coloured glimmer- 

 ing, be a consequence of the greater dispersion of the colours, 

 and consequent weakening of the last deep red, occasioned by 

 the lead it contains, or of any other dullness in the interior, I 

 could not discover; perhaps both causes may act in conjunc- 

 tion. That every dullness inside the glass does not produce 

 the same result may be seen from Experiment 22 ; which 

 was made with a prism full of small blebs, and which gave a 

 very faint spectrum, and where the maximum of heat still fell 

 into the red. 



If we now further consider the effects of flint-glass and 

 crown-glass, it will appear as a main result, — That the limits of 

 the prismatic spectrum are not confined, as usually assumed, to 

 the termination of the bright colours, 'ischere they cease "with a 

 kind of faintly coloured edge ,- but that the spectrum must extend 

 further, at least to that point lahere the greatest eJJ'ect is pro- 

 duced, although no colon?; or at least a very faint one, should be 

 perceivable there to the naked eye. That light is there, cannot 

 be doubted ; and in fact, as we must now add, visible light ; 

 since several philosophers, without respect to the increas- 

 ing confusion of language, still think themselves obliged 

 3 L 2 to 



