SOLAR ECLIPSE, MAY «, 1883. 123 



wlierc •' *' ' 'represents the phase. Of the nature of this function we know little except in every 



A 



l)ractical case it is a continuous, non-perioilic function of the time, and that its value must depend 

 upon the physical characteristics and temperature of the source. For integrations which involve 

 very small changes in the value of the function, or in which it may be regarded as independent of 

 t, Fuiosnel's solution is accurate, but in cases where E is very great it fails. 



With our limited linowledge of the nature of f{x,t) it is perhaps hopeless to attempt by ana- 

 lytical methods alone to find the illumination within the geometrical shadow; but it is easy to see 

 tliat nuuh more light would fall there than would be the case if Fresnel's solution obtained, for it 

 is the assumed perfect regularity of the phenomenon which causes the integrals to vanish when the 

 lower limit is considerable. Nothing remains then but an appeal to expei'iment. This was done 

 by em|»h)ying a lens system of 5 inches diameter, with a lime light in the more distant principal 

 I'ocutj as an artificial sun. It is evident that near the axis of the system this would appear as au 

 intensely Inilliaiit disk, and, at a distance of 50 feet, of about i degree iu diameter. This was 

 hidden to the eye by a blackened metal disk, 1 inch in diameter, placed at such a distance as to 

 be slightly greater in angular diameter. It was suspended by three fine wires at the center of a 

 ring for supi)ort. Under these circumstances the black disk was seen surrounded by a vivid ring 

 of light, even when of considerably greater diameter than the artificial sun. This ring was of 

 intense brilliancy next the disk and faded rapidly outwards. Its height was from 10' to 30' and, 

 as a whole, had a most striking resemblance to the solar corona except in its great regularity. 

 That most of the light was derived from the edge of the luminous disk was proved by the eflect of 

 stopping off the center of the lens. This was done by au assistant, and it was found impossible to 

 recognize whether the stop, cutting off perhaps 90 per cent, of the light, was in place or not as far 

 as the light ring was concerned. If, howevei", the stop was moved towards one side of the lens, as 

 its apijroach to that side became very close, the bright ring seen around the black disk began to 

 grow less bright on the corresponding side until there was a well-prouounced "rift." It is clear 

 from the experiment that if the brightness of any portion of the edge of the artificial- sun could be 

 greatly increased there would have appeared the complement of this rift, namely, a "streamer." 



This experiment, which is easily tried and highly interesting, is by no means conclusive, 

 though very suggestive. The conditions, with the dimensions readily attainable to practice, much 

 more nearly meet those of Fresnel's case than those of a total eclipse. Still, it must be regarded as 

 a valuable guide in si)eculating on the nature of the corona. The first evident objection to apply- 

 ing the principle to the explanation of the corona is that it would seem to imply a rapid change in 

 the extent of the corona during the eclipse, which according to it ought first to show much greater 

 brightness and extent on the eastern side, then, near third contact, on the western side. There 

 .seems to be no ocular evidence of such change and but little photographic evidence. But there 

 are considerations which greatly lessen the force of this objection. First, the change can oidy be 

 rapid during a very brief interval after sec^ond contact and before third contact, so that it might 

 readily escape observers whose attention is not specially directed to it; while, on the other hand, 

 the cauiera cannot gi\'e evidence of rapid changes since, in the case of the outer corona, the ex- 

 posure cannot be very short. Secondly, there is another reason why such changes should not be 

 very great, namely, the rapid diminution in brightness of the sun's disk in apjiroaching the limb. 

 From this it follows that at second or third contact the sun could be replaced, as far as the i)he- 

 uomenon in (piestiou is concerned, by a uniforndy bright disk of somewhat smaller dimensions, so 

 that the ratio of the angular distances of the limbs, east and west, will not be indefinitely great 

 or small. 



