rs, the same pair of reference-stars being used on each occasion.
The displacements caused by parallax were thus in a sense doubled, since
the star to which the planet seemed approximated in the northern
hemisphere, showed as if slightly removed from it in the southern, and
_vice versa_. As the planet pursued its course, fresh star-couples came
into play, during the weeks that the favourable period lasted. In these
determinations, only heliometers were employed. Dr. Elkin, of Yale
college, co-operated throughout, and the heliometers of Dresden,
Gottingen, Bamberg, and Leipzig, shared in the work, while Dr. Auwers of
Berlin was Sir David Gill's personal coadjutor at the Cape. Voluminous
data were collected; meridian observations of the stars of reference for
Victoria occupied twenty-one establishments during four months; the
direct work of triangulation kept four heliometers in almost exclusive
use for the best part of a year; and the ensuing toilsome computations,
carried out during three years at the Cape Observatory, filled two bulky
tomes[779] with their details. Gill's final result, published in 1897,
was a parallax of 8.802", equivalent to a solar distance of 92,874,000;
and it was qualified by a probable error so small that the value might
well have been accepted as definitive but for an unlooked-for discovery.
The minor planet Eros, detected August 14, 1898, was found to pursue a
course rendering it an almost ideal intermediary in solar
parallax-determinations. Once in thirty years, it comes within fifteen
million miles of the earth; and although the next of these choice epochs
must be awaited for some decades, an opposition too favourable to be
neglected occurred in 1900. At an International Conference, accordingly,
held at Paris in July of that year, a plan of photographic operations
was concerted between the representatives of no less than 58
observatories.[780] Its primary object was to secure a large stock of
negatives showing the planet with the comparison-stars along the route
traversed by it from October, 1900, to March, 1901,[781] and this at
least was successfully attained. Their measurement will in due time
educe the apparent displacements of the moving object as viewed
simultaneously from remote parts of the earth; and the upshot should be
a solar parallax adequate in accuracy to the exigent demands of the
twentieth century.

The second of the nineteenth-century pair of Venus-transits was looked
forwa