streamers. To attack the problem with any hope of perfect success, especially when the full extent of the outer corona is desired, I would propose the plan of Horatius encountering the three Curiatii, separate the enemy. If the maximum extension of coronal streamers be desired, the outer corona must be photographed independently. The inner corona should be just on the margin of the field. Its position, though out of sight, should be well defined; otherwise all future measurements would become impossible. The exposure for a Seed's 26 may range between 30 and 50 seconds, whereas the full eclipse could be captured in 3 to 5 seconds. With skillful manipulation, I have no doubt that a negative can be obtained giving greater extension to the streamers than photography has yet been able to accomplish. Now we may turn to the question of exposure. No doubt we should always aim at a correct exposure. In developing a welltimed plate we may expect a minimum of difficulties: but how seldom in out-door photography will an exact exposure be obtained! In photographing a bright object, the difference of a second of time may either under or over-expose a plate. In the hands of a skillful operator this accidental time will be counteracted by proper manipulation. Many factors enter into the calculation of a correct exposure, viz: the altitude of the sun, the condition of the atmosphere, the light-ratio of the lens, the size of the stop, the focal length of the objective, the proper illumination of the object to be photographed, etc. Atmospheric disturbances are, perhaps, the most important factors to deal with in celestial photography. When the altitude of the sun is small and the atmosphere is disturbed, it is impossible to obtain a great extension of coronal rays. But let us suppose that all the conditions are most favorable; that the objective is well corrected for spherical aberration and chemical rays; totality happens in summer; the sun is near culmination in an Italian sky; still there remain serious difficulties which will always make a total eclipse the most difficult picture to be taken in the whole range of photography. I am inclined to favor long exposures whenever very faint details in the shadows are to be photographed. A long exposure does not necessarily imply an over-exposure; but often leads to it. I should not fear an over-exposure so long as I could prevent a fog. I believe that more delicate details may be obtained from over-exposure than otherwise. The negative may be flat, wanting in density, but this defect may be corrected later on by intensifying the negative. Overexposure, however, has its limits. It does not follow, by any means, that if an exposure of one minute, to coronal rays, gives a fair amount of extension to the streamers, that doubling the time a greater extension will be obtained. No, you have crossed the limit of over-exposure, and a fogged plate is the inevitable consequence. One of the negatives of Prof. PRITCHETT'S party, taken at Norman, Cal., January, 1889, received an exposure of 30 seconds. Although the polar filaments are almost blurred from over-exposure, and the Eastern motion of the moon has left its trace on the inner corona, still this cliché gives the greatest extension to the outer This, however, must be noticed, that bright objects, such as a cloud near the sun, will hardly admit of over-exposure on account of the intense diffused light in the field. Very sensitive plates, such as a Seed's 26 are unfit for such work; I should rather use a Seed's 23, with a very small stop and a quick shutter; or, still better, a Vogel's Orthochromatic plate. corona. Another point of vital importance to be attended to, is the necessity of shielding the sensitive plate from all foreign light. A Seed's 26 will be affected by any light whatsoever; even by the so-called non-actinic red light of the dark room. This may not be noticed on the high lights of the picture, but the soft and delicate shades in the deep shadows will most certainly suffer, if the negative be exposed uselessly to the non-actinic light of the dark room. I would recommend to insert the plates into their carriers and to begin the process of development in total darkness. Allow me to cite an example which I think is to the point. Of all the different parties, which took part in the California eclipse of 1889, Prof. PICKERING'S station at Willows was certainly the best equipped. The long experience and skill of this able Astronomer, raised the expectation of all, and naturally invited us to look for the best results from that quarter, but a fatal oversight prevented the expected results. Instead of having a shutter attached to his 13-inch telescope, a plain board was placed 9 inches from the objective, in order to mask the sensitive plate between the different exposures. The diffused coronal light which entered the instrument sideways. and affected the plates as soon as the slide was removed, was sufficient to give a slight fog to his negatives, and thus marred the beauty of his best eclipse pictures. Another consideration which may have escaped the notice of experimenters, is that the exposure is greatly modified by the developer. The same exposure may give all the result of an under-timed picture with one developer, and of an over-timed negative with another. An experiment of last winter will make my meaning clear. With a six-inch objective, 92-inch focus, I made four exposures on the crescent moon. All the plates used were Seed's 26; time one-fourth of a second. In developing plate no. I, I used a cold pyrogallic solution. Result—an under-timed plate; no details. The second plate was developed with a normal solution of Hydroquinone. Result, -more details, but still under-timed. Plate no. III, was brought out by Eikonogen. Result,-perfect details, appearance of a well. timed negative. For Plate no. IV, I used the same Pyrogallic developer used on Plate no. I, only it was heated to 130 degrees. Result, an over-exposed negative which had to be retarded with K Br. This proved to be the best negative, on account of the richness of its details. I should add, however, that all brands of plates will not stand this high temperature. In winter, Seed's and Cramer's plates will give beautiful results, with this warm treatment, especially for instantaneous work. The gelatine film becomes very soft, hence great care is required in all subsequent washings. By way of recapitulation, I may recommend the following, in order to obtain greater extension of coronal streamers. I-To use Orthochromatic plates. I consider Vogel's Eoside of silver the best, when fresh, or Seed's 26, when developed with a warm pyro developer in winter. II-The greatest precaution to guard from all foreign light. corona. IV-Long exposures to secure the extension of the outer corona. V-Photographing each wing separately, and keeping the brighter part of the eclipse out of the field. ST. CHARLES, MISSOURI, October, 1890. PHYS THE KENWOOD PHYSICAL OBSERVATORY. BY GEORGE E. HALE. At the request of Prof. HOLDEN, I am glad to write the following account of our new observatory in Chicago. The special nature of the work for which it is designed may give this paper an interest it would not otherwise possess. Two In the summer of 1888 we erected a small brick building for spectroscopic purposes on the corner of Drexel Boulevard and 46th Street. It contains a general laboratory, "slit-room," "gratingroom," and photographic dark-room. With the exception of the laboratory, the walls and ceilings of all the rooms are painted dead black, and light-tight shutters and curtains on the few windows assure almost perfect darkness. A concave grating, mounted as described by Prof. ROWLAND, is the principal instrument. The grating-room contains three brick piers situated at the vertices of a right triangle, the right angle being at the north-east corner of the room. heavy wooden beams are supported on the piers, and form the sides about the right angle. To each beam an adjustable steel rail is bolted, and these would meet, if produced, at the apex of the right angle. At this point the slit is placed. It is very accurately made, with jaws of glass-hardened steel, and is capable of rotation about a horizontal axis, by means of a tangent screw, in order to set it exactly parallel with the lines of the grating. The partition between the slit room and grating-room is so built that it comes directly behind the slit, the light entering the grating-room through a short tube passing through the partition and screwed to the back of the slitplate. Thus the slit is entirely without the grating-room, and any light source can be used before it without the least danger of fogging the photographic plate on which the spectrum is received. A carriage moving on the north and south rail carries the grating, and is connected with a carriage on the east and west rail by a girder about ten feet long. On the latter carriage is held an eye-piece for observing the spectrum, or a plate for photographing it. The concave grating was ruled by Prof. ROWLAND. The ruled surface is 35% inches long and 13% inches wide, and has a radius of curvature of about 10 feet. It contains 14,438 lines to the inch, thus making available the whole of the first two orders of spectra, and a portion of the third and fourth. For work on very faint lightsources, such as the electric discharge in rarefied gases, a second grating of only 5 feet radius is used, of course with a short girder. This grating gives almost all of its light in one of the first spectra, and is exceptionally useful for the purpose mentioned. The gratingholder can be rotated about either one of three axes, and, when in use, the face of the grating is at right angles to the girder. With the long focus grating, photographic plates 10 inches long can be used, by bending them to a curvature of about 5 feet. Different regions of any spectrum are brought on to the plate by moving the carriage. along the track. The focus is the same at all points, and the photographed spectrum is normal. A building near by contains the dynamo from which electric currents are obtained. It is a 70-volt WESTON machine, and is driven by a gas-engine of 6-horse-power. A set of 35 JULIEN storage cells can also be used when desired. The current is led into the slit room to a specially constructed arc lamp or a large induction coil, suitable resistance being interposed in each case. The image of the arc or spark is thrown on the slit by a quartz lens, and the spectra are readily photographed edge to edge with the solar spectrum. Sunlight is thrown on the slit by a heliostat, placed on a pier far enough to the north of the building to be out of the shadow. Suitable absorbing solutions before the slit serve to cut out the overlapping spectra. When photographic enlargements of spectra are required, they are made in the slit-room by the aid of camera lenses. A GEISSLER pump is employed with various forms of vacuum tubes for the study of gaseous spectra. In such cases the required exposure is much reduced by using the short focus grating. The rooms are so connected by double doors that it is possible to pass through them all without disturbing any plates which may happen to be exposed. They are lighted by incandescent lamps when necessary. A considerable amount of experimental work was carried on in this building during the summer months of 1888 and 1889. The capacity of the apparatus was fully tested by a long study of the solar spectrum, notably in the region of H and K. The arc spectra of many metals were also photographed, and the reversals investigated. But my absence from the city during a large portion of each year, made it impossible to conduct any continued research, and the apparatus has always been dismounted in the winter. In the summer of 1889 Mr. BRASHEAR built for me a large telespectroscope, which was used last winter in solar work at the Harvard College Observatory. This instrument has already been described, as well as the research conducted with it. (Technology Quarterly, No. 4, 1890.) A frame of strongly braced steel tubing carries the two telescopes, which make with each other a constant angle of 25°. The objectives are exactly alike, about 34 inches clear aperture and 421⁄2 inches focus, and are made of Jena glass. The grating is of the same size as the large concave grating described above, but is, of course, plane instead of concave. The jaws of the slit move equally in both directions from the center, and the whole |