On his return from France in the autumn of 1904, Wood moved all of his spectroscopic apparatus from his room on the ground floor to a small room in the tower of the laboratory that supported the dome of the Johns Hopkins astronomical telescope. Here he was able to have sunlight all day, as he was above the shadow of McCoy Hall, which was just across the street. The light of the electric arc which he had used for the study of the fluorescence during the previous year was not intense enough for the complete solution of his problem, but with sunlight he had high hopes of bringing out some new and interesting phenomena.
The American Academy of Arts and Sciences in Boston had given him a liberal grant from the Rumford Fund, which enabled him to construct a large and powerful spectrograph, with three huge prisms of dense flint glass five inches square and large achromatic lenses. These were mounted on a rigid metal frame of steel rods and aluminum, together with the slit and plate holder, and the whole was placed in a cheap, unpainted pine box, the shape of a grand piano. This design was characteristic of all his subsequent apparatus. He didn’t care how the apparatus looked on the outside, provided that the concealed optical parts were of the highest perfection. A later instrument he called his tombstone spectroscope, as its base was a slab from a cemetery. His object now was to amplify the results that he had obtained the previous year with Moore, namely, that, as the color of the light employed in exciting the fluorescence of the sodium vapor was slowly altered from the blue to yellowish green, the region of maximum intensity of the fluorescent band spectrum was shifted in the opposite direction, i.e., from yellow to green. The simple steel tube used in the previous work had to be recharged with sodium after an operation of an hour or so, owing to the comparatively small amount of metal that could be used and its rapid distillation to the colder parts of the tube, where it hung in festoons of black spongy material from the wall of the tube. Cleaning the tube was a hazardous operation, as the metal had formed an explosive compound either with the hydrogen or nitrogen or both. The glass windows were removed and the tube was stood on end against the wall of the laboratory in the back yard. A pailful of water was then poured from a second-story window down into the open end of the tube. This operation caused a rapid series of terrific explosions, with great belches of yellow flame, which sometimes brought policemen into the vicinity in search of the lawbreaker who was discharging firearms within the city limits.
Says Wood:
To avoid these frequent annoyances to the police, I arranged a hollow three-inch drum of steel, which fitted snugly into the long steel tube, and had two small apertures, one for the entrance and the other for the exit of the concentrated beam of colored light. This drum was three-quarters filled with fragments of sodium before its introduction into the large tube, and the tube could now be operated for a hundred hours before it was necessary to clean it out and refill it.
Sunlight, reflected from a heliostat on the window sill, was focused on the slit of the monochromator, and the colored radiation emerging from the second slit was focused on the front aperture of the drum, forming a colored spot of fluorescence where the rays entered the sodium vapor that was streaming out through the aperture. An image of this spot was formed on the slit of my new spectrograph, by means of a mirror and condensing lens. The spectrum could be observed visually or photographed, and the changes in the distribution of the intensity as the prisms of the monochromator were turned, altering the color of the light thrown into the tube, were very remarkable. With blue-light excitation of the fluorescence, the spectrum consisted of two or three narrow yellow bands, but as the color was changed to blue-green and then to green, new bands appeared in the fluorescence spectrum which widened rapidly on the green side until it met the narrow band of the exciting light and finally extended beyond it on the short wave-length side, a certain and very striking exception to Stokes’s law of fluorescence. The very crude theories of fluorescence which had been developed up to this time broke down completely in this case, indicating clearly that the physical processes concerned in the case of fluorescence were enormously more complicated than had been assumed.
And now came the greatest discovery of all. Observing the spread of the fluorescent spectrum from yellow through green to blue, as the color of the exciting ray was slowly altered from blue to green, I thought I saw a slight trace of a fluttering movement in the broader bands in the green. Narrowing the slits of the monochromator, which made the exciting ray more nearly of a single wave length (purer color), to the point at which I could just barely see the fluorescent spectrum, I threw a black cloth over my head and the spectrograph, and saw to my amazement, in place of the more or less continuous spectrum of bands, a series of sharp narrow lines at regular intervals, like the divisions on a measuring stick. On slowly turning the screw of the monochromator which rotated the prisms and altered the color of the exciting light, the lines of the fluorescent spectrum appeared to be in rapid motion, vibrating to the right and left, the appearance being not unlike that of moonlight on rippling water.
The sight of a spectrum with lines wavering to and fro with an undulating motion was as unbelievable as would be the sight of a foot rule on which the divisions of the scale were moving about in an irregular manner. Closer observation showed that the lines were not actually moving, but were disappearing in one place and appearing in another. I now had a method of causing the appearance of various groups of widely separated lines in a complicated band spectrum of thousands of closely spaced lines, a method that would greatly simplify the study of these very imperfectly understood spectra. To cite another analogy from acoustics, it was as if one trying to form a theory of the structure of a piano by listening to the noise produced by slamming a board down over the entire keyboard had suddenly found that the keys could be struck successively or in groups[7].
Wood was jubilant as Archimedes! He had no bathtub to jump out of. But he wanted to celebrate, and did, in a manner which terrified the eyewitnesses and stampeded a mule team.
Never before (Wood says) had anyone ever seen the lines of a spectrum jiggle about in this manner. And just then a black cloud came over the sun, and there was a rumble of distant thunder. The storm came up rapidly, and in a few minutes it was so dark that lights went on in many of the rooms across the street. Then the rain came, a cloudburst, and soon a muddy torrent which stretched from curb to curb was running down the hill. A crowd had sought shelter from the rain in the wide porch of McCoy Hall across the street, and I suddenly thought of a splendid way of celebrating my discovery.
The thunder crashes were following the lightning at intervals of only a second or two. From my bottle of metallic sodium I picked out a lump the size of a small hen’s egg and, opening the window, waited for another lightning flash. A Negro was driving a cart up the grade against the wind and muddy stream half a block away, lashing a decrepit horse. Presently a terrifying flash occurred. I threw the metallic ball down, and it struck in the middle of the street, going off with a gigantic yellow flame and a bang that were coincident with the retarded detonation of the thunder. The crowd in the doorway fled precipitately into the building, and the Negro hurriedly turned his horse and cart and dashed off full speed down the hill, looking back over his shoulder at the volcano of yellow fire that was floating on the surface of the water and pursuing him.
7
This was the starting point of some of his most important investigations in what Wood termed “resonance spectra,” for which the theoretical physicists could not even suggest a plausible explanation. With the advent of the Bohr theory of energy levels in atoms and molecules, their explanation was at once evident, and as Wood had foreseen, their study was of the utmost importance in connection with the newer theory of band spectra.