rings, at ▲ 3388 and A 3456, which have been noted by several observers since 1898.

Great care was taken in focusing the objective-prism spectrographs, by the following method:

A parabolic silver-on-glass reflector, diameter 10 inches and focal length 10 feet 2 inches, by PETITDIDIER, was mounted in a horizontal tube of seasoned wood, and pointed to the mirror of the cœlostat in such a position that the rays from a bright star (parallel light) would cover the parabolic mirror and be brought to a focus on its axis. This focal point was determined very accurately by photography. A spectroscope slit was then mounted with its surface exactly in this focus. An iron electric spark from electrodes a few inches out from the slit was focused on the slit. The cone of rays passing through the slit completely covered the parabolic mirror. The spectrograph to be focused was mounted with its first prism just to one side of the slit so that the (parallel) rays from the mirror would cover the prism, and the ray of desired wave-length be brought to the center of the photographic plate. The usual methods of focusing the plate were then applicable. The entire process worked well and was exceedingly convenient.

OBJECTIVE GRATING SPECTROGRAPH.

In charge of Dr. KOBB.

This instrument, shown just beyond the ultra-violet spectrograph in the illustration, received its light from the Lick Observatory cœlostat. A plane Rowland grating, 14,438 lines to the inch, set for the third order, received the light, and returned it through a (visual) camera-lens 2%-inch aperture, 201⁄2-inch focal length, to a Cramer's isochromatic plate. The region of the green coronal ring at A 5303 was central in the field. A yellow-green color-screen was fastened immediately in front of the plate to cut out the overlapping spectrum. The purpose of the exposure, extending throughout the total phase, was to record the green ring, in order to determine whether the layer giving rise to its light was uniformly distributed around the Sun, or not. Owing partly to the clouds and partly to the light-consuming properties of the optical train, the image of the ring secured is exceedingly faint, and not much can be said from it as to the law of distribution, but there is little. doubt that the ring is of quite irregular intensity.

The distribution at other recent eclipses, at times of sun-spot minimum, was exceedingly irregular.

THREE-PRISM SLIT SPECTROGRAPH.

In charge of Mr. PALMER.

The purpose of this instrument was to determine the wavelength of the green coronal line near à 5303. It was mounted on a clock-driven polar axis, pointed directly toward the Sun, and received its light from an image-lens (visual) of aperture 11⁄2 inches and focus 101⁄2 inches. The slit-jaws were curved to a radius of 3.30 inches to make the recorded spectrum lines straight for ease and accuracy in measurement. It was placed east and west across the Sun's image. The collimator-lens (visual) was of 2%-inch aperture and 201⁄2 inches focus. The three extra dense flint prisms had refracting angles of 60°, refracting edge 2%, 24, and 24 inches, respectively, and length of faces 35% inches. The deviation for A 5303 was 169° 36'. The triple camera-lens (visual) had aperture 2% inches and focal length 201⁄2 inches. A movable diaphragm in front of the slit had six holes so arranged that the coronal spectrum from the east limb of the Sun would have a sky spectrum on either side of it for comparison, and similarly for the coronal spectrum on the west limb.

The images of the green line on the plate, exposed throughout totality, are likewise very faint, and capable only of approximate measurement.

SINGLE-PRISM SPECTROGRAPH.

In charge of Mr. PALMER.

This instrument was mounted beside the one just described, and received its light in the same way. Its purpose was to record the general spectrum of the corona.

The image of the Sun on the slit was formed by a lens of 15-inch aperture and 13 inches focus. The collimator-lens was of 2 inches aperture and 32 inches focus, corrected for Hy. The prism was of Jena glass No. 0.102, angle 63° 27', refracting edge 24 inches, length of face 4.55 inches. The cameralens was of aperture 2 1/16 inches and focal length 12 inches, corrected for HS. The region between Hy and H8 occupied the center of the field.

The exposure continued throughout the total phase. The

spectrum is strongly recorded. That of the inner corona shows no Fraunhofer dark lines, whereas they are strong in the spectrum of the outer corona. They are also shown, less strongly, on the area occupied by the Moon. It is thus clear that our own atmosphere, or perhaps the clouds, diffused the light to a considerable extent, and the results are more difficult to interpret on that account.

THE POLARIGRAPHS AND PHOTOMETER.

In charge of Professor MCKINNEY, assisted by Professor ROCASOLANO, Mrs. CAMPBELL, Mrs. PERRINE, and Srs. BARBAJOSSA and BLASCO.

The observations of the polarized light in the corona of 1901, with a double-image prism, were very successful. Owing to the dispersion of the prism, however, and to the small scale of the images, it was not possible to measure the intensities with the necessary accuracy for determining reliable values of the polarization. In addition to the double-image camera, which was again used, new apparatus was designed by Mr. PERRINE for polarization observations at the recent eclipse. Three cameras composed this apparatus. Two of these had planeglass reflectors in front of the objectives to serve as analyzers, while the third camera was used to secure an unpolarized image of the corona as a standard of comparison. The aperture of this direct camera was reduced so that the image obtained with it would be of approximately the same intensity as an (unpolarized) image with the other two cameras, after reflection from the plane-glass surfaces. The plane-glass analyzers were set at the angle of maximum polarization. Their principal axes were adjusted, one parallel to a north-and-south line and the other to an east-and-west line, through the corona. this way polarization was observed along four different radii of the corona. The three cameras have focal lengths of 50 inches, which give images of the corona 21⁄2 times the diameter of that obtained with the double-image camera. The four polarigraphs were mounted on one polar axis.

In

The performance of these cameras was highly satisfactory. They yielded sharp images, which are of sufficient size to permit quite accurate determinations of intensity to be made.

Four sets of negatives were secured with these instruments, with exposures of 1, 4, 20, and 115 seconds. Of these the I-second and 115-second series are unsuitable for the best

results, owing partly to the diffused light from the clouds, which affected the sky background differently in the different cameras. This effect was enhanced in the case of the first series by some of the slides having been drawn unnecessarily early.

The 4-second and 20-second series are well suited for accurate determinations of the amount of polarized light in the inner-middle and middle corona. No numerical results. have yet been obtained, as special apparatus is necessary for the photometric measurement of these negatives. The polarization is well marked, however.

An effort was made to compare the brightness of the corona with that of the full Moon by impressing a series of standard squares near the ends of two dry plates, with the Moon as the source of light, and then exposing the central portion of each plate to the light of the corona in a suitable camera, without any lens. This camera was arranged to admit the light from a circular area of sky 4° in diameter, the corona being in the center.

Two exposures were made during the eclipse of 14 seconds. and 50 seconds, respectively.

It will be necessary to determine the effect of the clouds and of the sky background before any definite result can be arrived at.

The negatives were developed at the station on the four days and nights following the eclipse, simultaneously with the dissecting and packing of the instruments. It was a satisfaction to find so large a proportion of the photographs of such excellent quality, notwithstanding the clouds. A cloud of any kind over the Sun at an eclipse looks pretty thick to those who have been preparing for fourteen months to observe the event!

The photographs and instruments, carefully packed for the long journey home, were shipped from Alhama on the evening of September 3d to Barcelona and thence by steamer sailing directly to New York. Farewell visits to the officials. of Alhama, the settling of accounts, and the closing of a voluminous correspondence terminated our eclipse duties abroad.