plates have given poor results in daylight, and Professor Nipher recommends Cramer's 'crown' plate.
The new star in Perseus, which has now waned in the sky, and in the memory of most people, is still an object of discussion among astronomers. Our readers will remember Professor Newcomb's recent article on variable stars and the difficulties in the way of accounting for their periodicity. In the extreme case of new stars the difficulty is greatest. The theories of an outburst from the molten interior and of collision might account for the appearance of the star, but do not explain its rapid waning, nor are they in accord with spectroscopic determinations. Professor Seeliger's theory that a dark star passes through a swarm of meteors is the most satisfactory form of hypotheses, but leaves room for the ingenious suggestion, recently made by the great astronomer, M. Janssen, before the Paris Academy of Sciences. He points out that the apparent absence of oxygen from the sun may be due to its existence in some dissociated condition that the spectroscope would not reveal. This condition may be owing to a very high temperature, and when this becomes low enough to allow oxygen to assume its common form, and so to unite with hydrogen, there would ensue, as a result of the combustion, a great increase in heat and light, which would account for the brilliancy of a new star. The rapid decrease in brilliancy which follows would be accounted for by the formation of an atmosphere of vapor, which would serve as a gradually increasing obstacle to radiation from the star. A corollary of M. Janssen's supposition is that our own sun may at any time reach this transition point for oxygen and blaze out into a fury of heat and light that would scorch all life off the face of the earth. It is, however, a pleasant feature of solar catastrophes that astronomical time is measured by millions of years.
One of the most interesting total eclipses of the sun, which the present century furnishes, will occur on May 18, 1001. The maximum duration of totality, which will be about six and a half minutes, is rarely surpassed. This will give exceptional opportunity, provided the sky is clear, for work of any kind, photographic or visual. The region of totality is, however, inconveniently remote, and the weather conditions, which usually prevail at the stations which will be occupied, are not of the best. The shadow begins off the east coast of Africa, a short distance to the southwest of Madagascar, sweeps northeasterly over the Indian Ocean, and crosses Central Sumatra, Southern Borneo and New Guinea, and a few smaller islands. To visit the track of the eclipse from New York, therefore, one must journey half way around the earth, and it matters little, so far as distance is concerned, whether one starts east or west. In spite of the distance, observations will be undertaken by a number of American and European astronomers. In this country, the Yerkes, Lick, Columbia, Amherst and Naval Observatories and the Massachusetts Institute of Technology will be represented by skilled observers. Under the auspices of the Royal and Royal Astronomical Societies, English observers will be stationed at Mauritius, and Padang, on the west coast of Sumatra. At Padang the eclipsed sun will be only 21° from the zenith, and the duration of totality about six and a half minutes. At Mauritius, the chances for a clear sky are much greater than at any other station, but the duration of the total phase is only three and a half minutes. On this account, nearly all the American and European observers are planning to visit Sumatra. The observations may have an added value from the fact that the eclipse occurs near the time of minimum sun-spot activity.
The chief part of the work in this, as in other recent eclipses, will doubtless be photographic. Owing to the enormous advantages which photo-
