Diffuse Reflexion of Light on Matt Surfaces. 211 



believes to be the only existing matt substance, was observed 

 at two different places in the spectrum which, according to 

 Helmholtz, ^Physiologische Optik } (p. 317), are complementary 

 colours. The same applies to plaster of Paris. The mixed 

 waves were measured in the same way as with ultramarine by 

 removing the prism of the photometer. The accuracy was 

 only 6 per cent. ; but to these last-mentioned measurements 

 no especial importance was attached ; they were only made 

 for reasons which will be subsequently shown. 



Even the best mirrors scatter light, which has been ob- 

 served by the photometer. From the results it is evident 

 that this scattered light obeys perfectly different laws from 

 those of matt substances. 



No regular Reflexion from the Plates. — The following table 

 gives the absolute differences of the observations from +-£ 

 and — e : 



English red | + 0-0024 



Chromate of potash 



Zinc green 



Carbonate of r red part of spectrum 



magnesia. I blue part of spectrum 



Plaster of r red part of spectrum 



Paris. i blue part of spectrum 



i=60, 



e = 20. 



6=20. 



i=80, 

 e=40. 



j + 0-0024 



+ 0058 



-0076 



+ '0028 



-0152 



+0382 



4- -0007 



0000 



+0435 



•oooo 



4-0082* 



+ 0327 



4- -0050 



4-0201 



+0168 



4- -0050 



+0092 



+0183 



4- -0026 



+0058 



+0146 



i 



Taking even 



the angle of 80° still into account the 

 mean variation is only 0121, that is, 2*8 per cent. This 

 being very near (08 per cent.) the limits of accuracy of the 

 indications, we must admit that both the light intensities are 

 the same. This is a strong argument for the superiority o£ 

 my plates. 



Influence of Colour. — Comparing the results we find that 

 in nearly all cases for a constant e the values for varying 

 angles i are greater for red than for yellow, and these again 

 greater than for green. On the other hand the values for i = 

 constant and e varying are almost the same for all colours. 

 Hence the colour seems to be of greater influence for the 

 angle of incidence than for the angles of emission. Mes- 

 serschmitt found the first result also, and tried to explain it 

 by means of Fresnefs observation, that the limit-angle, from 

 which the regular reflexion begins, varies for different 



P2 



