August U, 1885.] 



SCIENCE, 



183 



moved from the other half of the plate. The 

 telescope was now so placed that the sun should be 

 hidden behind a paper disk, fixed at about twenty- 

 feet distant. A diaphragm of one centimeter aper- 

 ture was placed in front of the lens, and an exposure 

 of four seconds given to the sky. On development, 

 half of the plate, except where cut by the image 

 of the disk, was found uniformly darkened. On tlie 

 other half were four images of the sun, two of 

 which were lighter, and one darker, than the sky. 

 The third image of eight seconds exposure was of 

 exactly the same darkness as the sky; and it was 

 accordingly shown, that since the diaphragm used 

 with the sky was about four thousand times larger, 

 the sun was about two thousand times as bright, 

 photographically, as the sky in its immediate vicin- 

 ity. A number of plates were taken on different 

 days, when the sky seemed perfectly clear, and the 

 results indicated that the number varied in general 

 between a thousand and four thousand. Owing 

 to the diffraction produced by the small diaphragm 

 used in photographing the sun, which rendered the 

 image 1.6 times larger than it really should be, all 

 these figures must be multiplied by 1.6. 



Comparisons were then made in a similar manner 

 between the sky near the sun, and the full moon, the 

 latter taken with the full aperture of the lens, 3.65 

 centimeters, and the former with an aperture of .204 

 centimeter. Under these circumstances, with ex- 

 posures of fifteen seconds, the moon and sky dark- 

 ened the plate to about an equal amount. The 

 result of a number of experiments indicated that the 

 sky in the immediate vicinity of the sun was of about 

 four hundred times the intrinsic brilliancy of the 

 full moon. The ratio of the sky to the sun on this 

 same day was fifteen hundred, so that the light of 

 the moon was to that of the sun as one to six hun- 

 dred thousand. In some experiments which I made 

 in 1879,^ I found the visual ratio was one to three 

 hundred and fifty thousand. On account of the ex- 

 treme blueness of the sun, it was to be expected that 

 the photographic ratio should be somewhat higher 

 than the visual one. 



I next tried comparing directly the light of the sun 

 and moon on the same plate, in order, if possible, to 

 get a check on my results. The results, however, 

 were unsatisfactory, the ratio coming out as 1 to 

 300,000, or only one-half the former amount. Owing 

 to the diflBculties of the experiment, this discrepancy 

 may very well be referred to inaccuracies of the pho- 

 tographic plate, and changes in the sun's and moon's 

 light during the course of the experiments. In all 

 the results with regard to the sun, it must be remem- 

 bered that the figures must be multiplied by 1.6, on 

 account of diffraction. The two ratios, then, of the 

 light of the moon to that of the sun, stand as 1 to 960,- 

 000, and 1 to 480,000; and of these, I think, in connec- 

 tion with my visual result, the former is the more 

 correct figure. The moon at the time of these 

 observations, June 26, 12 m., had an altitude of 29°, 

 when the atmospheric absorption would amount to 



1 ProcffJlinoH of the American academy of arts and sciences, 

 1880, p. 246. 



about twenty per cent.^ Making this correction, we 

 have the photographic ratio of the moon to the sun, 

 as 1 to 760,000, or about twice as great as that to the 

 eye. This is, of course, only an approximate result, 

 as only very few observations were made, and as it 

 was entirely outside the course of our inquiry. 



Keturning, then, to our original subject, we found 

 the sky near the sun 400 times as bright as the full 

 moon. Correcting for atmospheric absorption, this 

 figure becomes 320 times. But we found before, that 

 in order to detect the contrast between a coronal ray 

 and a neighboring rift, the light of the sky must not 

 exceed that of the full moon. It therefore seems 

 that even in the clearest weather the reflected light 

 of the atmosphere is 300 times too strong to obtain 

 the faintest visible image of the true coronal rays. 



In connection with these experiments, I took a few 

 photographs of the sun with my improved apparatus. 

 In order to still further diminish the reflection of the 

 light from the surfaces of the lens, I so placed the 

 telescope that the sun was almost completely hidden 

 behind the high steeple of a neighboring church. A 

 vast improvement in the results was at once obtained. 

 The sun stood out sharply defined on a perfectly 

 uniform background of blue sky. There was not the 

 slightest trace of a fringe either where the steeple 

 crossed the disk, or where the sky came in contact 

 with the solar limb. The day was beautifully clear, 

 and at six in the afternoon some more photographs 

 were taken; but now, although the steeple was as 

 clear as ever, all around the limb of the sun appeared 

 the atmospheric halo, extending out in all directions, 

 and gradually growing fainter as it receded from the 

 sun. We may, therefore, in general, say, that with 

 properly constructed apparatus, in perfectly clear 

 weather, no halo whatever appears around the sun. 

 It is only in slightly hazy weather, or as the sun ap- 

 proaches the horizon, that the appearances are pro- 

 duced which have been elsewhere described. 



In brief, the result of my researches would seem 

 to indicate, 1°; that without a total eclipse it ought 

 to be impossible to photograph the solar corona, 2°; 

 having tried, I have failed to photograph the corona, 

 but have obtained the result which theory indicated. 



Wm. H. Pickeeing. 



STEIN EN'S EXPLORATIONS OF THE 

 XINGU. 



Dr. Karl vo^ Steinen" has recently made a most 

 interesting report of his explorations in the Matto 

 Grosso, — the immense region, more than four times as 

 large as France, which occupies a large part of central 

 and western Brazil, and is hardly known to geogra- 

 phers except in the most imperfect manner. It is 

 divided by great rivers, of which the Madeira, the 

 Tapajos, the Xingu, the Araguaya, and the Tocan- 

 tins flow northward, and the Paraguay flows south- 

 ward. It is watered by innumerable streams which 

 unite with these rivers, along whose banks live thou- 



1 Annals Harvard college observatory, vol. xiv. p. 62. 



