is got rid of, a mere shed serving all purposes of protection
required for the "coude."

The desirability of some such change as that which M. Loewy has realised
has been felt by others. Professor Pickering sketched, in 1881, a plan
for fixing large refractors in a permanently horizontal position, and
reflecting into them, by means of a shifting mirror, the objects desired
to be observed.[1646] The observations for his photometric catalogues
are, in fact, made with a "broken transit," in which the line of sight
remains permanently horizontal, whatever the altitude of the star
examined. Sir Howard Grubb, moreover, set up, in 1882, a kind of
siderostat at the Crawford Observatory, Cork. In a paper read before the
Royal Society, January 21, 1884, he proposed to carry out the principle
on a more extended scale;[1647] and shortly afterwards undertook its
application to a telescope 18 inches in aperture for the Armagh
Observatory.[1648] The chief honours, however, remain to the Paris
inventor. None of the prognosticated causes of failure have proved
effective. The loss of light from the double reflection is
insignificant. The menaced deformation of images is, through the
exquisite skill of the MM. Henry in producing plane mirrors of all but
absolute perfection, quite imperceptible. The definition was admitted to
be singularly good. Sir David Gill stated in 1884 that he had never
measured a double star so easily as he did Gamma Leonis by its
means.[1649] Sir Norman Lockyer pronounced it to be "one of the
instruments of the future"; and the principle of its construction was
immediately adopted by the directors of the Besancon and Algiers
Observatories, as well as for a 17-inch telescope destined for a new
observatory at Buenos Ayres. At Paris, it has since been carried out on
a larger scale. A "coude," of 23-1/2 inches aperture and 62 feet focal
length was in 1890 installed at the National Observatory, and has served
M. Loewy for his ingenious studies on refraction and aberration--above
all, for taking the magnificent plates of his lunar atlas. The "bent"
form is capable of being, but has not yet been, adapted to
reflectors.[1650]

The "coelostat," in the form given to it by Professor Turner, has
proved an invaluable adjunct to eclipse-equipments. It consists
essentially of a mirror rotating in forty-eight hours on an axis in its
own plane, and parallel to the earth's axis. In the field of a telescope
kept rigidly pointed tow