194. 



KNOWLEDGE 



[October 2, 1898. 



suggested by Sir H. Grubb, and indeed the true lightning lines are 

 easily distinguished under the magnifying power, as they haye a centi*al 

 liue with distinct penumbra, wliieli the lamp lines have not. . . . 

 "Belieye me, dear sir, yours faithfully, 



" W. D. Hemphill " 



Fig. 2. — Photograph of setting sun taken by Mr. Gloycr in the mouth ot May, 1 893 



A close examination of the photograph shows that the 

 bright spots upon the sea are joined by lines which, after 

 following a very irregular course, lead to the 

 gas-lamp. The circular halo round the gas- 

 lamp is evidently caused by reflection of light 

 from the back of the plate, and shows that the 

 flame of the gas must have been very bright. 

 It seems to have been sufficiently bright to trace 

 a line on the sensitive plate as the camera was 

 moved, producing a larger over-exposed patch 

 wherever the camera momentarily stopped. The 

 camera seems to have been stopped and held 

 steady for a considerable time in the position 

 where the image of the lamp is brightest — 

 probably one of the lightning flashes occurred 

 during this interval — but there is a double 

 image of the horizon and of other distant objects 

 which it seems probable were produced at the in- 

 stant when the landscape was lit up by the two 

 flashes of lightning shown on the left hand of 

 the picture, the camera being pointed in slightly 

 different directions at the instant of each flash. 



The circular halo which frequently appears 

 round the photographic image of small bright 

 objects has caused much speculation to observant 



termine the cause of similar halos which are sometimes seen 

 round the bright images of prominences, and found that the 

 diameter of the halo varied with the thickness of the glass 

 plate on which the photograph is taken. When a brilliant 

 point of light is photographed, and the image 

 is over-exposed, its photographic image is sur- 

 rounded by a brilliant halo, the inner edge of 

 which is comparatively sharp while the outer 

 edge is very soft and nebulous. On measuring 

 the thickness of the plates used, and the dia- 

 meter of the circle corresponding to the inner 

 sharp edge of the halo, I found that the inner 

 edge of the halo corresponded to the place where 

 hght from the luminous image was reflected at 

 the ci'itical angle from the back of the plate. 

 Diu'ing the exposure the parts of the photo- 

 graphic fllm which are being acted upon are 

 brilliantly illuminated, and light must be dis- 

 persed from the illuminated parts of the film in 

 all directions within the thickness of the plate. 

 The light which falls nearly perpendicularly 

 upon the back surface of the plate in great part 

 emerges and is lost in the space behind the 

 plate, but a part is reflected, and on reaching the 

 film on the front surface pioduces, if it is suffi- 

 ciently intense, a photographic effect. Of the 

 dispersed light which falls more obliquely on the 

 back of the plate, a larger proportion is reflected, but it is 

 distributed over an area which increases with the distance 



Fia. 3.— Photograph of setting sun taken by Mr. Gloyer, 13th June, 1893. 



photographers, and has been a fruitful source of theories with 



«©• 3(r 20* w* 



wao* JO* vf 



Fio. 4. — Diagram showing the intensity of light reflected from 

 the back of a photograpliic jilate. 



regard to photographic action. Some years ago, when study- 

 ing eclipse photographs, I made a series of experiments to de- 



from the illuminated part of the film, so that the intensity 

 of the photographic action of the reflected light on the 

 film at first decreases till the angle of total reflection at 

 the back of the plate is approached, when the intensity of 

 the reflected light rapidly increases. 



In order to further verify this theory, I calculated ac- 

 cording to Fresnel's law the intensity of the hght which 

 would be reflected from the back of the plate, and would 

 reach the film on the front surface at various distances 

 from the luminous point, on the assumption tiiat the photo- 

 graphic plate was made of crown glass, with an index 

 j«. =: 1-54, and that light would be uniformly dispersed 

 in all directions within the plate from the luminous part 

 of the film during the exposure. Fig. 4 is a curve, the 

 height of which corresponds to the intensity of the light, 

 on the above suppositions, reaching the front surface of 

 the plate after reflection at various angles from the back 



