'The Transactions of the Bombay Geographical Society. From January 1863 to December 1864. (Edited by the Secretary.) Volume XVII.' [‎106] (135/524)

cvx
PROCEEDINGS OF THE
Mr. Burgess said Dr. Haines had remarked on the former p a
that the detection of errors of the kind pointed out by Mr. Broun tend*
ed to shake our confidence in the whole of the observations, but this
instance seemed a much more remarkable one of the way i n which
errors had been perpetuated in the observations for no less a period
than sixteen years; and though, as Mr. Broun had remarked, we are
told in other volumes that in the course of that time the results had
been again verified, it was to be looked upon as indicative of a want of
independent mind in the conduct of the Observatory, that during this
long period the mistake had not been detected. When no similar
result had ever since been noticed elsewhere, it was strange that this
solitary instance had not been suspected, and the whole question refer
red to some one competent to deal with it, or that the data were not
re-examined with the corrected temperature co-efficient. It seemed still
more remarkable, that when the effect was said to have been verified at
later dates, the difference of the results arrived at by using a co-effi
cient of 2*0, instead of the older and more erroneous one of 3*2 small
scale divisions, should not have been noted as accounting for a very
large proportion of this supposed effect of the surrounding air. Then
the consequence of this hypothesis was a serious one, for all the quanti
ties to which it has been applied in these 1 5 volumes, are vitiated by it
as well as by the error Mr. Broun had pointed out in the temperature
co-efficient. It was greatly to be regretted that for so long a time the
observations had been corrected by a value of this co-efficient, which a
rough examination of the observations themselves showed to be inaccu
rate to the extent of at least a fifth of its amount. It was of importance,
he thought, that both the temperature and unit co-efficients should be
determined with the utmost possible accuracy before they were applied
to the simple leadings. The long standing of these mistakes reminded
im of another, no curious, if of much less consequence; he referred
to the piinted scale-readings of the bifilar magnetometer; by some
accident or another, at the commencement the divisions for tens had
been read as units and those for units as tenths ; and this had been
continued ever since—thus the variations appeared to be only one-tenth
o t e amount at Bombay, which they might be expected." The most
cuisoiy inspection of a volume of observations from any other Observa-
tory s ou have made this patent, yet strange to say it seemed never
een noticed, and the same mistake was continued from the