A SOUTHERN OBSERVATORY. a les 
heavenly bodies could be depended upon to within ten minutes; with 
the new Repsold heliometer the error of a single observation is less 
than one-tenth of a second of arc. So that accuracy has been increased, 
‘in the course of three and a half centuries, some six thousand times. 
At what cost of patience and expenditure of the counted moments of 
individual human lives, as the fruit of what illuminations of genius, 
throes of invention, failures, and disappointments in some quarters, 
compensatory triumphs in others, can never rightly be told. The prog- 
ress achieved was by “leaps and bounds;” it must henceforth be by 
slow and painful foot lengths, as the limit of possible accuracy is brought 
imperceptibly nearer. It is not likely that the astronomical data of 
three and a half centuries hence will be six thousand times more ac- 
curate than those at our disposal. 
The heliometer is, of all others, the instrument best adapted for the 
work (exceedingly simple in principle, yet delicate to an almost incon- 
ceivable degree in the details of its execution) of determining stellar 
parallaxes. The diameter of the earth’s orbit affords a base line 
186,000 miles in length, from opposite extremities ofwvhich—that is, 
at opposite seasons of the year—the distances between the object to 
be examined and two ‘‘comparison stars” are measured. The infinites- 
imal alternate shift of the star nearest the earth to and from those with 
which it is compared (assumed with little risk of error, to be indefi- 
nitely remote) is called its “‘ parallax.” From its angular amount the 
distance in miles of the star from the earth can be at once derived. 
The minuteness of this little parallactic see-saw is difficult to be real- 
ized by those unpracticed in such matters, A displacement of one sec- 
ond on the sphere is equivalent to a shifting across the width of a 
human hair placed 70 feet from the eye. But no known star has so 
large a parallax as one second, which is as much as to say that no 
known star is so near to us as 200,000 times the distance of the sun. 
Positive results might, under these circumstances, well have been de- 
spaired of; yet they have, in a number of cases, been attained, and 
form the surest groundwork so far provided for investigations into the 
mechanism of the skies. 
Dr. Gill’s observations for stellar parallax were begun at the Cape 
July 5, 1881, with the Dunecht heliometer, of which he had become the 
possessor by private purchase from the Earl of Crawford. He had asa 
coadjutor Dr. W. L. Elkin, who is now in very effective charge, at Yale 
College, of the only heliometer yet erected on any part of the American 
continent. Nine stars in all were measured, of which two gave no in- 
dications of possessing any sensible parallax. Both, remarkably enough, 
are brilliant stars of the first magnitude—Canopus and Beta Centauri— 
which, to shine as they do, from unfathomable depths of space, must be 
objects of astounding splendor. Canopus, especially, can not emit less, 
and may emit a great deal more, than fifteen hundred times the light 
of our sun, unless, indeed, Dr. Elkin’s “comparison stars” shofild turn 
