eclipses, is due to the existence of meteor streams. It is also undoubtedly true that several of the meteor systems encountered by our earth in her journey round the sun have the vast dimensions mentioned by Professor Abbe. Indeed, he far underrates the dimensions of the August and November meteor systems, each of which must be measured in length by hundreds of millions of miles, not by mere millions. But it is absolutely impossible that any of the meteor systems traversed by our earth, or any meteor systems of no greater degree of richness, should present the appearance of streamers surrounding the sun, like those in our figure above. So far as the two systems specially mentioned by Professor Abbe are concerned, inasmuch as we know the exact shape and position of the orbits along which the meteors forming these systems travel, we can determine the exact position which the meteoric streams occupy in the heavens at any moment; and most certainly neither of them on July 29th last occupied the position of the two beams shown across the sun in our figure. The August system was the one which at the time passed nearest to the sun's place on the sky, but it did not come within several degrees of the sun. The November system did not even cross the part of the sky where the sun was. These two systems, therefore, could not possibly be connected in any way with the two streams, of whatever nature, which produced the rays intersecting exactly at the sun.
But there is a more general objection to the theory that such meteor systems may explain coronal streamers seen during total eclipses of the sun. If such streams could be seen when situated beyond the sun, they would be seen far better when opposite the sun on the dark background of the midnight sky. Take, for instance, the November meteors. We know that the flight of meteors, some 2,000,000,000 of miles long, which the earth traversed in November, 1866, 1867, 1868, 1869, 1870, and 1871, is now nearing the remotest part of the long orbit of the November system, many millions of miles beyond the path of Uranus. We know that at midnight in winter the richest part of that system lies due south, at an elevation varying from 30° to 50° above the horizon. There, illuminated fully by the sun, though at a great distance from him, it ought to be far better seen than a similar system lying beyond the sun and visible only through the light of the brightest part of the corona. But no one has ever, on the darkest and clearest night and under the most favorable atmospheric conditions, even suspected the existence of the faintest possible light where the heart of the November system is really situated. Much less, then, could such a system be seen during total eclipse (if so situated as to lie athwart the sun). Systems less rich than the November system (the richest known to us) would have still less chance of being discerned.
If, then, we are to account for the radial streamers seen by Professor Abbe, and also seen during many other total eclipses, though to a less distance, by the meteoric theory, we must consider meteor systems
