A history of any department of practical astronomy, written from the point of view of a mechanician, could hardly fail to be of interest. Among the curious experiences at Harvard in the line of photometry is one which illustrates this point, and, at the same time, indicates the refinements in observation which are resorted to, and demonstrates one of the utilities of the photometric method. [Illustration: HARVARD OBSERVING STATION IN PERU.] In 1877 announcement was made of the discovery at Washington of two satellites of the planet Mars. The Harvard telescope being applied they were after a little effort descried as two faint points of light, showing no visible disks. To ascertain the diameter of each satellite might therefore seem impossible: but it was done, approximately, by the photometric method. The mechanical problem was to reduce the light of the planet as seen in the telescope to an equality with the light of one of the satellites as thus seen. Five or six different mechanics were employed to drill in a piece of metal a hole, making a true circle, and small enough to produce the equality sought for by sufficiently diminishing the light of the planet. It may be remarked that one of those who succeeded best had already, for his own purposes, managed to drill a hole, lengthwise, through a fine cambric needle, making a steel tube of it. What he made for Prof. Pickering was a hole in a steel plate, the diameter of which was one eighteen hundredth (1-1800) of an inch. It was so nearly circular that the various diameters, including errors of measurement, only differed one one hundred thousandth (1-100,000) of an inch. Other mechanical devices were resorted to for corroboration, and the results reached were that the diameter of one of the satellites is about six miles, and that of the other about seven miles. They are the smallest known in the solar system. The availability of the spectroscope in astronomy had early been appreciated by the profession. In experiments in this line it had been found that a classification of the nebulæ might be made upon the basis of their spectra. In 1880 the study was carried a stage further at Harvard in ascertaining by the spectroscope that certain faint objects, which, by direct vision, had been judged to be stars, are in fact nebulæ. In 1881, it was found that the spectroscope is serviceable in the discovery of variable stars. Thus incited, a new instrument was imported from London, but it did not prove satisfactory.