It is appropriate that a commerative volume marking an important epoch in the life of our friend and collegaue Professor Stratton should contain a communication, however slight, in the field to which he has devoted a lifetime of fruitful and constructive research, the field of solar and stellar spectroscopy. In addition, it is fitting that it should deal with work of a pioneering character carried on under somewhat difficult physical conditions, because Professor Stratton is no stranger to such situations, having conducted solar eclipse expeditions to many of the waste places of the Earth. The subject which I have selected has to do with the development of solar research during the early years of the Mount Wilson Observatory in Southern California. It is familiar ground for Stratton, who in later years has visited it occasionally, and has always maintained a deep interest in its growth and progress.
In 1903, Dr. George E. Hale, Director of the Yerkes Observatory, spent a winter in Southern California. A distinguished solar investigator and an ardent ``sun worshipper'', as he called himself, he enjoyed the many sunny days of the semi-tropical climante and often turned his eyes toward the range of mountains to the northward which bounded the coastal plain. The most accessible of these, although the accessibility was limited to two foot-trails passable by donkeys and mules in an air-line from the small city of Pasadena. Hale made several trips to Mount Wilson and observed the Sun with a 4-inch portable telescope, each time with increasing enthusiasm over the observing conditions.
On his return to the Yerkes Observatory Hale made application to the recently established Carnegie Institution of Washington for a grant to bring a large solar telescope to Mount Wilson upon an expeditionary basis. The grant was made and with three members of the Yerkes staff to accompany him Hale started on his new venture.
The Snow Telescope, as the instrument was named after its donor, was wholly of the reflecting type and consisted of three mirrors. The first and largest of these, a flat mirror 30 in. in diameter, received the direct sunlight. It was mounted in a cell with its axis pointed toward the north pole, and rotated by clockwork to follow the apparent course of the Sun. A second flat mirror, somewhat smaller in size, placed to the south of the large mirror and slightly above it, intercepted the beam of reflceted sunlight and sent it northward through a long house to an image-forming concave mirror. This third mirror had a focal length of 60 ft and formed an image of the Sun 6 1/4 in. in diameter. Rotation of this mirror made it possible to throw the image upon an auxillary instrument within the house.
Fortunately the telescope had no very massive parts, since methods of transportation were primitive, and the limits of size and weight were fixed by what the pack animals could carry over nine miles of twisting mountain trails. The limits of weight were about 200 lb for the powerful mules, and 150 lb or less for the donkeys, or burros, to give them their accepted western name. All these animals were highly tempermental, and their characteristics were doubtless responsible for the development of the now rapidly vanishing group of men of remarkable skill and lurid vocabulary who drove them. It is interesting to realize that in the Snow Telescope building, of steel construction and 180 ft long, no single piece of the framework exceeds 8 ft in length because of these restrictions upon transportation. The problem of the very few pieces of the mounting of the telescope which exceeded the standard limits was solved by Hale though the invention of a narrow truck, with a length of 10 ft and a tread of 20 in., which was drawn by mule and steered at both ends. This picturesque vehicle was used until the trail was widened into a road to provide for the transportation of the larger telescopes of later years.
Construction of the large stone piers to support the telescope and auxillary instruments, and of the building to house them, was completed in 1904-5. The building itself was of unusual construction, the inner wall consisting of continuous sheets of canvas, and the outer wall of canvas louvers which permitted circulation of air but shut out direct sunlight. The problem of reducing the injurous effect upon the definition of the solar image of warm currents of air rising from bare ground heated by the Sun led to several experiments. The growth of brush and grass over bare areas near the telescope was encouraged, artificial covering with light white cloth was tried to a limited extent, and some tests were made of the effect of stirring the air inside the telescope building. Probably the most remarkable experiment was that of hoisting an observer and a 4-in. telescope to a height of some 50 ft in the air in a large pine tree, and comparing the quality of the Sun's image observed from this height with that of ground level. It was these experiments which led Hale two or three years later to design the tower telescope, in which the sunlight incident upon the first mirror is received high above the ground and is reflected downward through a shielded vertical tube. A marked improvement was observed in the quality of the image, and this form of telescope has been adpoted widely by other observatories.
The auxiliary instruments used with the Snow Telescope were in general three in number. The first was a high-speed photographic shutter which provided direct photographs of the Sun, showing in detail the structure of sunspots and other features of the surface. The second was a spectrograph for analysing the Sun's light and studying the behavior of the thousands of spectral lines of the elements which compose the Sun's atmosphere. The third was a spectroheliograph, an instrument invented and used almost exclusively by Hale. With this ingenious device the distribution of the white-hot clouds of individual gases over the surface of the Sun may be photographed and their forms analyzed.
Two major investigations were undertaken with these instruments, both of them dealing with the nature of sunspots. One was with the large spectrograph, the other with the spectroheliograph. Although quite separate in their beginnings their final results have united to provide an adequate explanation of most of the observed characteristics of sunspots, and at least a partial insight into the phenomena of their origin.
The first of these investigations was a photographic study of the spectrum of sunspots. It had been known to astronomers for a number of years that the spectrum of a sunspot differs from that of the surface of the Sun outside a spot in several respects. The whole spectrum of the spot is fainter, as might be expected from its comparative darkness, but in addition there are two important differences in the spectral lines. First, many lines are strengthened in spots as compared with the disk, and others, fewer in number, are weakened. Secondly, nearly all the lines are considerably widened in spots, some even appearing double. Few, if any, photographic observations, however, had been made, due to the fact that spectrographs of adequate power and size could not be attached to moving telescopes, and forms like the Snow Telescope had not come into general use except for solar eclipses.
The first photographs of the sunspot spectrum showed a wealth of detail which made it possible to catalogue many hundreds of lines according to their behavior. The most interesting feature was that all the lines of the same element did not behave alike. Some of the lines of iron, for example, were greatly strengthened in the spot as compared with the disk, others were little changed, and a few were weakened. These weakened lines were for the most part recognized as belonging to the class of ``enhanced'' lines, that is, lines which when investigated in laboratory sources are found to be greatly intensified in the spectrum of the electric spark as compared with the electric arc. For the lines which were strengthened, however, no such differentiation was available. This was long before the theory of ionization had provided a logical explanation of the classification of spectral lines according to energy levels in the atom. Hence the Mount Wilson work was of necessity almost wholly on an empirical basis.
As a working hypothesis, it was assumed that the temperature of a sunspot is lower than that of the general surface of the Sun. This seemed reasonable on several grounds, and it was soon proved to be correct through the discovery of molecular bands of titanium oxide in the yellow and red portions of the spectrum. The next step was to try to imitate the behavior of lines in the spot spectrum through observations in laboratory sources. A small spectroscopic laboratory was available on the mountain, and Professor Gale, of the Physics Department of the University of Chicago, who was spending a year at Mount Wilson, took an active part in the observations. The first element examined was iron which has a spectrum rich in lines showing wide differences of behaviour in sunspots. At first the procedure was to photograph the spectrum of a direct-current iron arc, varying the amount of current through as large as range as possible, but later the simple method was adopted of comparing the outer and cooler flame of the arc discharge with the inner and hotter core near the poles. In both cases considerable differences in temperature were attained.
An examination of the photographs showed conclusive results. Many lines were relatively strengthened in the outer flame or cooler portions of the arc, and many others were most intense in the hot central core. The correspondence with sunspots was striking. The lines of iron conspicuously stronger in spots than on the disk were those which were most relatively strengthened at low temperatures in the arc, while the lines which were unaffected or sometimes slightly weakened in spots were those which were prominent at high temperatures in the arc. The investigation was later extended to other elements prominent in spot spectra, such as titanium, vanadium, and chromium, with very similar results.
The products of this investigation have had some interesting applications. Not only did they solve many of the complexities of the sunspot spectrum, but they also provided a means for separating spectral lines into classes according to their behavior with temperature. When the theory of ionization was developed this temperature classification because of great importance, since it formed a starting point in the problem of analyzing the spectral lines of the various elements according to energy levels in their atoms, one of the great accomplishments of modern spectroscopy.
In later years, when the 60-in. reflecting telescope was in operation at Mount Wilson, the spectra of stars were examined in detail with a view to applying the results found in sunspots. Many stars have spectra closely resembling that of the Sun, and many others that of sunspots. In general the agreement was found to be close, but as more and more stars were examined some anomalies were discovered. A few spectral lines seemed to be abnormally strong or weak in spectra which otherwise were nearly identical. These differences proved to be associated with the intrinsic luminosities of the stars, or the total amount of light they emit. The luminosity of a star may be readily calculated from its apparent brightness, if its distance is known, merely by applying the inverse-square law of distance. At the time of this investigation relatively few accurate determinations of stellar distances were known. Enough were available, however, to establish a satisfactory correlation between the intensities of the anomalous lines in the spectra and the luminosities of the stars observed. Once such a correlation was established the process could be reversed, and the previously unknown distance of a star could be calculated from its apparent brightness and luminosity.
The final result of this application of the sunspot investigation was the discovery of a new method for determining the distances of stars, and for classifying them into groups according to luminosity and possible order of evolution. It forms an interesting illustration of the ramifications of researches in physical science and of the unforeseen consequences which may follow them.
Another discovery of extraordinary interest came out of the study of sunspots begun with the Snow Telescope. It had its origin in the use of the spectroheliograph. The earlier work with this instrument had consisted in the recording of the calcium flocculi, or clouds of calcium gas, over the surface of the Sun. Such clouds are especially numerous and dense in the vicinity of sunspots, and measurement of their areas furnishes an excellent index of the general state of solar activity. Their forms, however, are irregular, and often they overlie and obliterate the spots, due no doubt to the high level of the calcium gas in the Sun's atmosphere.
Hale had long wished to utilize spectral lines other than those of calcium in the spectroheliograph, more especially those of hydrogen, which is the most abundant element in the Sun and stars. The practical difficulties, however, were considerable at this time. The strongest and most suitable line of hydrogen lies in the red portion of the spectrum, and photographic plates sensitive to red light were not as yet available commercially. Still it was known that certain organic dyes increased sensitivity to red light, and that emulsions treated with these dyes, when dried rapidly, and exposed immediately, might give satisfactory results. So, the experiment was tried, the slit of the spectroheliograph being set on the α-line of hydrogen across a spot group near the center of the Sun.
The resulting negative had little to commend it from the photographic point of view, but scientifically it was of great interest. Two features characterized the clouds of hydrogen gas around the sunspots. In the first place they were dark against the solar background instead of being bright as were the calcium clouds. In the second place the hydrogen filament showed distinct evidence of vortical structure, with spiral arms centring around the centres of the spots. Later photographs indicated that over areas of intense solar activity portions of the hydrogen clouds might even become bright and be seen as the ``solar flares'' which affect the transmission of radio waves in the Earth's atmosphere.
It was the vortical structure around spots which interested Hale particularly. He remembered Rowland's observation that a rapidly rotating electrically charged plate produces a magnetic field; and Zeeman's discovery that a magnetic field splits up the spectral lines emitted by a source placed in the field and polarizes them, that is, separates the plane of vibration of the light-waves of the different components according as they are viewed along or across the lines of magnetic force. Hale remembered, too, the widening and occasional doubling of lines observed in the spectrum of sunspots. Could the vortical motion observed indicate rapid revolution within the spot-vortices of streams of electrically-charged corpuscles, corresponding to Rowland's charged plate, and thus producing a magnetic field? If so the field could be detected through the Zeeman effect and the polarization of the sunspot lines. The optical apparatus required to test for the effect was simple to construct, and on a day late in June, 1908, Hale examined a suitable spot with a large spectrograph and the 60-ft tower telescope, which had recently been completed. The results were definite and conclusive in showing that sunspots are the centres of strong magnetic fields, and thus provided the first clear evidence for the existence of magnetic forces outside the Earth.
With the discovery of the Zeeman effect in sunspots the second chief characteristic of their spectrum, the widening and occasional doubling of the lines, found a rational explanation. In later years there followed at Mount Wilson three discoveries of importance: the reversal and return to normal of the magnetic field in spots in the two hemispheres of the Sun in a period of twenty-three years; the existence of very strong magnetic fields in certain stars; and most recently the presence of changing fields of moderate strength over the surface of the Sun outside of spots. The impetus given to solar research by the early sunspot investigations has persisted throughout nearly half a century.
This brief outlines of some of the early solar research at Mount Wilson does not include any attempt to define the contribution of each individual to the investigations undertaken, except in the case of Dr. Hale. He was the leader at all times, and his insight and enthusiasm were the driving force and inspiration of those early years. Most of the research was carried on through the collaboration of two or more individuals, and the results were published jointly. These may be found in Volumes 1 and 2 of the Contributions of the Mount Wilson Observatory, and in the Astrophysical Journal for the years 1904-1908.