ture of the water changed, the slight expansion caused by the friction would have made the threads vary slightly. This would have caused the carriage that runs on it to vary slightly, and consequently the spaces between the grooves on the "grating" would vary, and render it useless for scientific purposes. The screw is turned by a solid steel wheel with seven hundred and fifty teeth on the ring, which is moved the space of one tooth at a time by an ingenious contrivance attached to the driving shaft. The screw having twenty threads, the carriage is moved one fifteen-thousandth part of an inch each time, thus making that many grooves to the inch on the metal "grating." The number of grooves may be regulated. "Gratings" have been made with forty-eight thousand grooves to the inch. By the strongest microscope made, the human eye could not see the lines if there were more than a hundred and twenty-five thousand to the inch. Prof. Rowland says he gets the best results from "gratings" with fifteen thousand grooves to the inch. The machine now in use at Johns Hopkins University is the third of the kind made. The first was completed years ago, and is still in use in the vault. The European, and especially the German universities, have tried repeatedly to make a machine of the kind, but have never succeeded. Hence all their best universities get the "gratings" for their spectroscopes from the machine at the Hopkins.
When a "grating" is completed, it is taken out, tested, packed in a handsome hardwood case, and sent to Mr. Brashear in Allegheny, Pa. This gentleman attends to the sale of the valuable "gratings," which cost from twenty to three hundred dollars. The proceeds are divided between Mr. Brashear and the university.
When these tiny grooves, cut with a diamond point on the polished metal plate, are completed and are perfect, the grating breaks up a ray of light into its various colors as a prism does. Some of the gratings produce "ghosts," and are then considered imperfect. Prof. Rowland deals with these "ghosts" of the spectrum in a recent article in the Astro-Physical Journal, of Chicago. He says: "A periodic displacement of one millionth of an inch in a grating will produce visible ghosts which are seen in the secondand are troublesome in the third. With very bright spectra these might even be seen in the first spectrum. An over-exposed photographic plate would readily bring them out."
With the concave grating Prof. Rowland has made an immense photographic map of the solar spectrum, and has determined a system of standard wave lengths which is now universally adopted. He is now having measured the wave length of every line of the solar spectrum and is determining the elements to which they all belong. This is a work of years, as is also the