gestions will be made. The instruments are divided into two classes, for visual and for photographic work. But in either case

the most important feature is the focal length or the size of the telescope. Since the photographic image of the corona will not bear magnifying without dispersing the available light, and thus blurring out the details of the picture, which is the most important feature to retain to the utmost, one can not use a short telescope and at the same time a magnifying eyepiece to enlarge the image by projection on a screen or on a photographic plate. The only alternative in order to get an image of large diameter is to use a long-focus lens. The effect of a difference of focus upon the image of the corona is well shown on Chart V, which gives a small corona (1) taken with a four-foot lens (Barnard), (2) with a fifteen-foot lens (Pickering), and (3) with a forty-foot lens (Schaeberle). The diameter is proportional to the focal length, but the difference of effect upon the details is very important. In the small picture the details of the corona near the sun are completely lost in the general light, while the coronal extensions from the middle latitudes are seen at a great distance from the sun-as much as one million miles; at the same time the polar rifts are distinctly marked, so that the pole or central line from which they bend is readily located. On the second picture the details of the polar rays are better brought out, but the extensions are shortened. In the third the region near the sun's edge has many interesting details very clearly defined, while all the extensions are gone. It is evident that each lens has its advantage, according to the details sought, and they ought all to be employed in the eclipse. The reproductions on paper by no means do justice to the original negatives, which make the distinctions even more pronounced than shown on Chart V.

Some amateur observers have telescopes but no mountings suitable for eclipse work, and many astronomical telescopes have good equatorial mountings at home which are yet unavailable in the field. for lack of proper foundations and supports. The ordinary telescope balanced near the center, with the eye end subject to all sorts of motions which may happen through jarring and rough handling in the hurry of shifting photographic plates, makes a very poor eclipse apparatus. All telescopes of any length should be held firmly at each end, so as to be perfectly steady, since the least vibration ruins a coronal picture devoted to delicate photographic effects. There are two ways of accomplishing this, and only two. Dispensing with an equatorial mounting, the lens must be set permanently on a base, and light reflected from a mirror must be utilized, which shall be concentrated on a plate also placed on a fixed base. This is the method employed by Schaeberle in Chili, April 16, 1893, to obtain No. III of Chart V. There is no objec

tion to it if an observer possesses a perfectly plane mirror, which it is very difficult and also expensive to obtain; if the reflecting mirror should be imperfect it would distort the image of the corona. The second method, lacking a good mirror, is to mount the longfocus lens in a tube and point it directly at the sun. A forty-foot lens was thus mounted at Cape Ledo for December 22, 1889, and its action was very satisfactory. Of course, it was a cumbersome arrangement, and could not be employed by a small party. The foundation for the mounting of the forty-foot tube consisted of two casks filled with stones and cement, and set firmly in the ground. These made two good piers, since the narrowing tops of the casks held the bed plates of the telescope as in a vise. A triangle, whose base was parallel to the earth's axis and having the telescope tube itself for the long side, was fitted with an extension rod for adjustment in altitude on the third side, and the whole was made to revolve on ball bearings. This triangular support was rotated by a side rod of adjustable length, whose end terminated in a sand piston working with a regulating valve. The sand flowed out steadily like an hour-glass, and dropped the tube, keeping it central on the sun. The image was made to follow accurately for twenty minutes without tremor, all the time holding the solar disk tangent to fixed lines. The principle of a revolving triangle and a short piston, taking the place of an expensive reflecting mirror with a delicate clockwork or one carrying a telescope balanced on its center but subject to jars and side motions, is an important assistance in field work on account of its ready adaptability to all sorts of observations. Since time is limited, it is necessary to provide all operations with automatic arrangements as far as possible, by using such an apparatus as that described. What can be applied successfully to a forty-foot lens can serve for shorter telescopes. In combination with spectroscopes, polariscopes, and special instruments for photographing, an immense amount of work can be compressed into the few seconds allowed by mounting them. all on such a movable frame as the triangle. The old-fashioned method of putting one observer to one telescope ought to be abandoned. Of course, for a rising sun during the forenoon a modification in the moving support must be employed. This should be such as to cause the objective of the telescope to rise from the ground toward the meridian, and it must be accomplished by attaching a heavy weight which in sinking draws the tube upward.

Of the two kinds of photographic plates, the wet and the dry, the dry plates are much more convenient in the field, and are good for certain purposes. The objections to them are that the rough granulations of the gelatin film sometimes overpower the fine de

tails of the corona itself; they burn out very near the disk by over-exposure while the faint outer extensions are being taken. Wet plates having the requisite quickness are harder to prepare,

CHART VI.-TRACKS OF THE EIGHT TOTAL ECLIPSES OF THE SUN SEEN ON THE NORTH
AMERICAN CONTINENT DURING THE NINETEENTH CENTURY.

but are smoother and hold the coronal rays better from the base to the outer edge, and there is always plenty of time to give the necessary exposure. The No. 26 Seed plate requires from 0.5 to

2.0 seconds only, generally one second being about right; the wet plate will take the corona in eight seconds or less. The best time of exposure should be tested on a bright star of about the second magnitude, by trial before the eclipse. There is no rule about the photographic focus, except to discover it by a series of exposures at different distances near the supposed point. Eclipse work is a practical matter, and many rough-and-ready methods must necessarily be admitted. A good lens in a wooden light-tight tube, supported at each end, having the motion of the sun, the photographic focal plane carefully determined, the time exposures very short, and, finally, exceedingly slow development of the picture after the eclipse these form the prime requisites. Expensive telescopes, clockwork on heavy iron piers, reflecting mirrors, and such like apparatus are not needed. Ingenuity in practical details, with great anxiety about the essential matter of the light itself, is what is needed for a successful eclipse expedition.

Those persons who have no telescope for viewing the sun, or camera for photographing it, can yet see the corona to great advantage by means of a good opera glass, and indeed this is really the most satisfactory way to thoroughly enjoy the spectacle. The object may be sketched on paper at once or from memory, and this picture may well be of value to astronomy.

The tracks of the eight North American eclipses seen since 1800 are shown on Chart VI. It is noted that three have paths very similarly located, and that five run in directions about parallel to one another, but almost at right angles to the first group. This comprises the eclipse of November 30, 1834, duration two minutes; August 7, 1869, two minutes and three quarters; July 29, 1878, two minutes and a half, which stretch from Alaska southeastward in a fan-shape to the South Atlantic coast. The second group contains the tracks June 16, 1806, four minutes and a half; July 17, 1860, three minutes; January 11, 1880 thirty-two seconds; January 1, 1889, two minutes and a quarter; and May 28, 1900, two minutes. These tracks all trend from southwest to northeast, and cross the North American continent in different. latitudes, that of May 28, 1900, being the most southerly and of rather short duration, lasting less than two minutes in the United States.

The path of the total eclipse of May 28, 1900, after leaving the United States, crosses the North Atlantic Ocean to Coimbra, Portugal, and continues over North Africa to its end at the Red Sea. Stations which are not situated on the path of the totality will see the sun partially eclipsed, in proportion to the distance of the locality from the central line to the northern or the southern lim