varies from one minute and twelve seconds near New Orleans to one minute and forty-four seconds near Norfolk, on the central line. These durations diminish from the maximum at the middle of the track to zero at the northern and southern limits of it, so that an observer must be stationed as near the central line as possible in order to see much of the eclipse. The population of several of the abovementioned cities is at present as follows: New Orleans, 242,000; Mobile, 31,000; Montgomery, 22,000; Columbus, 20,000; Atlanta, 66,000; Raleigh, 13,000; and Norfolk, 35,000. It is evident that with very little exertion more than 500,000 people can see this eclipse. It is most fortunate that the track passes near so many cities, because, with their facilities for the accommodation of visitors, many will be induced to undertake excursions with the purpose of taking in this rare sight, and a little enterprise on the part of railroads and transportation companies might easily increase the numbers. If people will go to a parade, yacht race, or an exposition, and consider themselves paid for their expenses, then surely they will find in this great spectacle of Nature not only an object of wonder and beauty, but also one of peculiar instruction in many important branches of science. All educators who can induce their pupils to make such an expedition will implant a love of astronomy in many impressionable minds which will become a source of pleasure to them for the rest of their lives.

Out of about seventy eclipses of the sun which have occurred somewhere in the world within the nineteenth century, there have been only eight total eclipses of more or less duration visible on the North American continent. The others happened in places often remote from civilization, and sometimes in entirely inaccessible localities, as over the ocean areas. The difficulty of transporting heavy baggage to the remote parts of Asia, Africa, or South America is such as to preclude all but a few scientists from any effort to observe eclipses. The writer was much impressed with the formidable nature of undertaking to establish eclipse stations in places which are distant from centers of population by his own experience on the West African Eclipse Expedition, sent out by the United States Government, for the eclipse of December 22, 1889, to Cape Ledo, on the west coast of Angola, about seventy miles south of St. Paul de Loanda. Nearly eight months were consumed in the course of the preparations at home and in the voyage out and back. The expedition, it should be said, however, went to Cape Town, South Africa, and halted also at St. Helena, Ascension Island, and Barbados for magnetic and gravity observations, so that all this time should not be charged to the eclipse proper. We sailed in the old frigate Pensacola, the companion

to Farragut's flagship, the Hartford, with Captain Yates. In earlier days Admiral Dewey commanded this ship, and the expedition was fitted out while he was in charge of the Bureau of Equipment at Washington. The same fine courtesy that has become so well known to his countrymen was at that time extended to all the members of the expedition.

The cloudiness along the track of the eclipse in the Southern States on the 28th of May, 1900, is evidently a matter of much importance not only for all astronomers, but for non-professional spectators. If it could be foretold, with the same precision as the astronomical data give the time and the place of the occurrence of the eclipse, that the day itself will be fair or cloudy, or that certain portions of the track will be clear while others will be obscured, it would be of great benefit. The cost of these scientific expeditions is very great, since it is necessary to transport many heavy and delicate pieces of apparatus into the field, including telescopes, spectroscopes, polariscopes, and photographic cameras, and set them up in exact position for the day of observation. The expedition to Cape Ledo, West Africa, in 1889, carried out a large amount of material, prepared it for work during the totality, and then entirely lost the sun during the critical moments by a temporary obscuring of the sky through local cloud formations. There had been some clouds at the station during the forenoons for several days preceding the eclipse, but the sky was usually clear and very favorable during the middle of the afternoons. The totality. came on at three o'clock, and photographs of the sun were taken at first contact about 1.30 P. M.; clouds thickened, however, and totality was entirely lost, while the sun came out again for the last contact at 4.30 P. M. This was a very trying experience, and of course could not have been avoided by any possible precautions. Some astronomers have thought that the advance of the moon's shadow is accompanied by a fall in temperature, and that cloudiness is more likely to be produced from this cause.

Soon after the West African eclipse Professor Todd, of Amherst College, proposed that more systematic observations be made for the probable state of the sky along eclipse tracks, with the view of at least selecting stations having the most favorable local conditions. The method was tried in Chili, April, 15, 1893, and in Japan, August 8, 1896, with some success. Heretofore the available meteorological records, which were originally taken for other general purposes, had been consulted, and some idea formed of the prevailing tendency to cloudy conditions. In accordance with the improved method, the United States Weather Bureau has been conducting special observations on the cloudiness occurring

from May 15th to June 15th in each of the three years 1897, 1898, and 1899, for the morning hours of the eclipse-between 8 a. M. and 9 A. M. A tabular form was sent through the local offices to such observers as were willing to act as volunteers in making these records, and their reports have been studied to discover how the cloudiness behaves along the eclipse track at that season of the year. Each of the three years gives substantially the same conclusion-namely, that there is a maximum of cloudiness near the Atlantic coast in Virginia, extending back into North Carolina, and also near the Gulf coast in Louisiana and in southern Mississippi, while there is a minimum of cloudiness in eastern Alabama and central Georgia. The following table will serve to make this plain:

The Prevailing Cloudiness of the Sky along the Eclipse Track.

The significance of these figures is shown by transferring them to a diagram, given on Chart II, which indicates the average cloudiness prevailing over the several States where they are crossed by the track. The marked depression in the middle portions, especially over Alabama and Georgia, indicates that the stations in these districts make a much better showing than those nearer the coast line. The reasons for this difference are probably many in number, but the chief feature is that the interior of this region, especially over the higher lands of the southern reaches of the Appalachian Mountains, which are from six hundred to one thousand feet above the sea level, is somewhat freer from the moisture. flowing inland from the ocean at that season of the year. The table shows also two divisions, one for the "general sky," wherein the relative cloudiness was noted in every portion of the visible sky, and for the "sky near the sun," where the observation was confined to the immediate vicinity of the sun. The two records agree almost exactly, except that the sky near the sun averages a little lower than the general sky. This indicates that although the sun will be seen in the morning hour of May 28th, when it is only from thirty to forty degrees above the horizon, yet this is not an unfavorable circumstance. The low altitude, on the other hand,