Intelligence and Miscellaneous Articles. 73 



the method of least squares. The derivate of that equation sires, 

 for each wave-length, the factor by which the intensities of the 

 prismatic spectrum must be multiplied to reduce them to those of 

 the normal spectrum. 



2. We have measured for wave-lengths comprised between 480 

 and 740, at intervals of 20 units, the rotations" which cause the 

 vanishing of the strife. 



3. The variations of the Carcel standard are corrected so as to 

 reduce the intensity to the constant value corresponding to the 

 consumption of 42 grams per hour ; for this we made use of 

 M. DeleuiPs automatic balance, which registers, during the time 

 of the experiments, the successive times required for burning 

 10 grams of oil. The solar light was diffused by a wooden screen 

 whitened with carbonate of magnesia. "We operated when the skv 

 was very clear, about noon. 



4. The prismatic curves have been reduced, as we have said, to 

 the normal spectrum ; and, lastly, the absolute maximum being 

 set off upon the curve, we have reduced the ordinates to those 

 corresponding to a maximum equal to 100. 



The curves obtained are tangents at their two extremities to the 

 axis of the wave-lengths ; they rise at first slowly, afterwards very 

 quickly in the vicinity of the maximum. For the lamp, the curve 

 is nearly symmetrical on both sides of the maximum. For the 

 sun, the ascent and descent are more rapid than for the lamp ; the 

 descent towards the violet is more rapid than the ascent on the 

 red side. 



The following results are deduced from the two most regular 

 series, selected from amongst numerous determinations : — 



Wave-length "40 720 700 680 660 640 620 600 



Illuminating- f Lamp 01 07 T6 57 14 28 52'5 94 

 powers ...\Sun 0'5 1-5 4 102 23 



Wave-length 580 560 540 520 500 480 



Illuminating- f Lamp 72-5 37"5 235 130 6-0 1-0 



powers ...\Sun... 625 98"5 ' 30"5 172 9-2 3-5 



The maximum, =100, corresponds, for the lamp, to the radiation 

 592, and for the sun to 564. 



The discussion of these results leads to conclusions which are 

 in accordance with the theory of the emission of the radiations 

 emitted by incandescent bodies *. 



The numbers obtained require a slight correction, on account of 

 the absorption exerted towards the violet by the material of the 

 prisms. We purpose to continue these determinations with prisms 

 exerting no appreciable absorption upon the visible radiations. — 

 Comjjtes Eendus de VAcademie des Sciences, Dec. 5, 1831, t. xciii. 

 pp. 959-961. 



* Journal de Physique, viii. p. 357. 

 Phil. Mag. S. 5. Vol. 13. No. 78. Jan. 1882. G 



