132 



SCIENCE, 



[Vol. YI., No. 132. 



inclination of the surfaces to one another at any point 

 was necessarily very slight; and this fact, combined 

 with tlie extreme thinness of the glass, rendered the 

 multiple internal reflections almost entirely harmless. 

 Five diaphragms were used between the lens and 

 plate, cutting off totally all internal reflection from 

 the tube. The same drop-shutter was used as before, 

 working just in front of the lens. 



Several observers have attempted to compare the 

 brilliancy of the corona as seen during a total solar 

 eclipse with that of the full moon. Thus Belli in 

 1842, and Wilson in 1860, find the corona brighter than 

 the moon; while Halley in 1715, Billerbeck in 1851, 

 and Curtis in 1869, find the moon brighter than the 

 corona.i Other observations by W. O. Eoss in 1870, 

 and by J. C. Smith in 1878,^ would indicate that the 

 corona was somewhat brighter than the full moon. 



A pliotograph of the solar corona in order to be of 

 any use whatever should show something more than 

 a thin uniform ring of light around the sun. It 

 should show some structure, some details of the rays 

 and rifts visible at the time of an eclipse. The only 

 observations which I have been able to find of the 

 intrinsic brightness of different portions of the corona, 

 were those by Professor John W. Langley in 1878.^ 

 He found that the corona at Y distance from the sun 

 had a brilliancy equal to six full moons, while at A! 

 distance it was only equal to .1 of a full moon. Un- 

 fortunately for our present purpose these observa- 

 tions were visual, and not photographic; but they 

 will give us an idea of the size of the figures with 

 which we are dealing. In order, then, for a photo- 

 graphic plate to show any of the irregularities of 

 detail in the corona, such as the contrast between a 

 ray and one of the neighboring rifts, it must be capa- 

 ble of showing contrasts of light which do not vary 

 from one another by more than about one-tenth the 

 intrinsic brilliancy of the full moon. 



A series of investigations was next instituted to 

 determine what excess of brilliancy it was necessary 

 for one surface to have over another, in order that 

 the contrast between them might be rendered per- 

 ceptible in a photograph. The difference of brilliancy 

 required by an acute eye amounts to between one 

 and two per cent. One sees it sometimes stated that 

 the camera is capable of distinguishing contrasts 

 which are quite invisible to the eye. This, I think, 

 is a serious error. If both sources of light are of 

 great brilliancy, though differing considerably, the 

 eye may not perceive a very great difference between 

 them; while a negative with a very short exposure 

 may show a very considerable difference. If, how- 

 ever, the eye be protected by colored glasses, the con- 

 trast will be as great as that presented by the negative. 

 But the question which occurs at present is not 

 where great differences of light exist, but where the 

 differences are very small, of only a few per cent. 

 The inferiority of the best gelatine plates to the hu- 

 man eye in this respect is very readily shown by an 



1 Memoirs of the Royal astronomical society, vol. xli. pp. 

 243-253. 



2 "Washington observations, 1876: Appendix iii., p, 387. 

 •* Washington observations, 1876: Appendix iii., p. 211. 



attempt to photograph distant mountains. It will be 

 found that it is perfectly impossible, even in the 

 clearest weather, to photograph at a much greater 

 distance than fifty or sixty miles. I recently, on a 

 very clear morning, made a mountain ascent with a 

 camera. Mountains over ninety miles distant were 

 readily recognized and distinguished; but, on point- 

 ing the camera in their direction, nothing over forty 

 miles distant could be photographed. There was no 

 question but tha't mountains at a much greater dis- 

 tance than ninety miles could have been seen, had 

 there been any high enough to be visible. Every 

 photographer who has visited mountain regions is 

 perfectly aware of the disabilities under which he 

 labors in this respect. 



Another illustration of the same thing is the im- 

 possibility of photographing the moon in the day- 

 time, when the sun is high above the horizon. 

 Although the moon may be perfectly distinct to the 

 eye, the negative shows no trace of it. This fact of 

 itself, I think, has a direct bearing on the question in 

 point. 



But in addition to these general facts it was 

 thought that some quantitative results would be de- 

 sirable. Besides the chloride plates which I had 

 been using, several well-known kinds of bromide 

 plates were tested at the same time. These were 

 selected with especial regard to the strong contrast 

 qualities which they were supposed to possess. The 

 plates tested were the Anthony chloride, the Carbutt 

 B, the Allen and Kowell, and the Stanley. Different 

 portions of the plate were exposed to a uniform illu- 

 mination for various times, and it was found that all 

 the plates gave about the same result, and that if the 

 division lines between the areas were very sharp, and 

 over an inch in length, as small a contrast as five per 

 cent could be detected; but if the division lines were 

 not over one-eighth of an inch in length, even if one 

 knew just where to look for them, it was impossible 

 to recognize a difference of less than ten per cent 

 upon the negative. As the coronal rays on the pho- 

 tograph would be less than one-eighth of an inch in 

 length in order to reach out beyond .3', ten per cent 

 was selected as the limit of contrast necessary to ob- 

 tain a satisfactory result. 



Since the light reflected by the corona at 3' dis- 

 tance from the sun is only .1 that of the full moon, 

 in order to distinguish between a coronal ray and a 

 neighboring rift at that distance, it is necessary that 

 the light reflected from the earth's atmosphere in 

 that region should not exceed in intrinsic brilliancy 

 that reflected by the moon itself. 



A series of observations was next made to deter- 

 mine the relative light of the sun and of the sky in 

 its immediate vicinity. The method employed was 

 as follows: Half of the photographic plate was 

 covered with thick yellow paper; a diaphragm of .016 

 centimeter in diameter was placed in front of the 

 lens, and four different exposures made to the sun on 

 different parts of the plate, lasting respectively for 

 two, four, eight, and sixteen seconds. The plate was 

 then taken into the dark room, and the exposed por- 

 tion protected by the yellow paper, which was re- 



