almost absolutely steady; the yearly rainfall
amounts to no more than three or four inches. No wonder, then, that the
"seeing" there is of the extraordinary excellence attested by Mr.
Pickering's observations. In the absence of bright moonlight, he tells
us,[1642] eleven Pleiades can always be counted; the Andromeda nebula
appears to the naked eye conspicuously bright, and larger than the full
moon; third magnitude stars have been followed to their disappearance at
the true horizon; the zodiacal light spans the heavens as a complete
arch, the "Gegenschein" forming a regular part of the scenery of the
heavens. Corresponding telescopic facilities are enjoyed. The chief
instrument at the station, a 13-inch equatoreal by Clark, shows the
fainter parts of the Orion nebula, photographed at Harvard College in
1887, by which the dimensions given to it in Bond's drawing are doubled;
stars are at times seen encircled by half a dozen immovable diffraction
rings, up to twelve of which have been counted round Alpha Centauri;
while on many occasions no available increase of magnifying power
availed to bring out any wavering in the limbs of the planets. Moreover,
the series of fine nights is nearly unbroken from March to November.

The facilities thus offered for continuous photographic research
rendered feasible Professor Bailey's amazing discovery of variable
star-clusters. They belong exclusively to the "globular" class, and the
peculiarity is most strikingly apparent in the groups known as
Omega Centauri, and Messier 3, 5, and 15. A large number of their minute
components run through perfectly definite cycles of change in periods
usually of a few hours.[1643] Altogether, about 500 "cluster-variables"
have been recorded since 1895. It should be mentioned that Mr. David
Packer and Dr. Common discerned, about 1890, some premonitory symptoms
of light-fluctuation among the crowded stars of Messier 5.[1644] With
the Bruce telescope, a photographic doublet 24 inches in diameter, a
store of 5,686 negatives was collected at Arequipa between 1896 and
1901. Some were exposed directly, others with the intervention of a
prism; and all are available for important purposes of detection or
investigation.

Vapours and air-currents do not alone embarrass the use of giant
telescopes. Mechanical difficulties also oppose a formidable barrier to
much further growth in size. But what seems a barrier often proves to be
only a fresh starting-point; and