Captain Abney on Photographic Irradiation. 51 



though the plane of maximum reflection will not pass through 

 the centre of the particle. The light reflected horizontally would 

 obey the ordinary laws of absorption, and the intensity be repre- 

 sented bv 



I' = If-", 

 where z is the distance along the film measured from the point 

 of incidence. 



If the relative effect on the sensitive film could be represented 

 by IT, where T is the length of exposure, then it would be easy 

 to calculate exactly the distance to which the irradiation from 

 this cause would extend. Experiment has shown that this simple 

 relation does not hold good, the practical results having been 

 given in the above-quoted paper. There it will be seen that the 

 shorter the exposure the less proportional irradiation there must 

 be. If, however, we take this relation to hold good, the effects 

 of irradiation in any example we may work will be in excess of 

 the truth. In the experiment (i) it was found that the limit 

 of irradiation by reflection with \ second's exposure was caused 

 by an intensity of sunlight ='0016, together with that caused 

 by the light from the sky. The sky-light would also equally 

 act on that part of the film which would be affected by the light 

 scattered parallel to the plate. The thickness of the film was # 004 

 inch ; hence /ti = 750 nearly for each inch. 



From other experiments taking images on opaque surfaces, it 

 was found that, with similar exposure and under similar con- 

 ditions, the extent of developable irradiation was nearly "005. 



By taking the mean intensity of the light passing through 

 the film, we can find approximately the proportion of light 

 which is scattered parallel to surfaces to that reflected from the 

 plate. Let m be ratio, 



•0016 = -3e- 750x, ° 05 xm; 

 m = -23. 



If the sun's image were photographed (say) 4 inches in dia- 

 meter with an effective aperture of 2 inches, instead of # 05 and 

 \\ inch respectively, it would be seen, calculating from the 

 above data, that with an exposure of -j-Jq" (supposing the simple 

 relation between I and T held good) no developable image could 

 be formed, much less could there be irradiation due to either 

 cause. Now it is well known that with the photoheliograph 

 under these conditions a good image of a clear sun is obtainable. 

 Evidently, then, the connexion between I and T is of a more 

 complicated nature. 



It will be noticed that a large difference in two intensities of 

 light (giving the same exposure on the same film) will make 

 but a small difference in the extent of irradiation, as the coeffi- 



E2 



