254 Profs. H. Rubens and R. W. Wood 



on 



Curve 2 shows a minimum at 46X D , a maximum at 85X D5 

 and a second minimum at 122*5 X D . Curve 3 has its first 

 minimum at 47"5 X D , a maximum at 90 X D , and a second 

 minimum at 128 X D . 1£ we calculate the mean wave- 

 length of the radiation complex, in both cases from the 

 first minimum, we obtain for the experimental series of 

 %• 2, Xi = 46 x 0-589 x 4 p = 1082 p. For fig. 3, 

 X/ = 47-5 X 3-589x4//, = 111-8 yu. The values calculated 

 from the positions of the first maxima in each case are 

 X 2 = 100ya and X 2 f =106yw, and from the second minima, 

 X 3 = 96'3^ and \ 3 ' = 100fjL. 



We thus see that we obtain slightly different values for 

 the mean wave-length according to which maximum or 

 minimum we use in the calculation. If we consider the 

 incidence angle as 6° instead of normal (owing to conver- 

 gence), these values are to be reduced by about 0'6 /bb. In 

 both cases the calculated wave-length decreases with in- 

 creasing " order " of the maxima and minima. The expla- 

 nation of this lies in the unsymmetrical character of the energy 

 curve. The first minimum gives us the value of the wave- 

 length corresponding to the centre of gravity of the energy 

 curve, while the following maxima and minima approach more 

 and more near to the wave-length of the maximum of the 

 energy curve. The energy curves in the two cases differ, not 

 only in the position of the wave-length of the centre of gravity, 

 but also in the position of the maximum and the degree of 

 asymmetry. It was foreseen that the asymmetry of the 

 energy curve would be greater for the thin plates than for 

 the thick ones, for in the latter case the radiation was 

 obliged to pass through an additional thickness of 10*6 mm. 

 of quartz, and the short waves would be more strongly 

 absorbed than the long ones, for which quartz is very trans- 

 parent. The rising slope of the energy curve will be 

 displaced towards the longer wave-lengths and made less 

 steep by increasing the thickness of the quartz, while the 

 descending slope will be influenced but little. 



It is possible to get an idea of the approximate form of the 

 energy curve of the radiation isolated by the quartz lenses, 

 by a trial and error method. We know the position of the 

 top of the curve approximately, and the terminus on the 

 short wave-length side (determined by the quartz absorp- 

 tion), and we can draw the curve on the long wave-length 

 side, by assuming that our source of radiation is non-selective, 

 and that the intensity decreases with the fourth power 

 of the wave-length. This curve we may now divide in 



