exceedingly; of course, the
longitude, by its Greenwich rate, was becoming considerably erroneous.
About this time, it should be remarked, the thermometer was varying from
65 deg. to 41 deg.. The greatest alteration we ever observed in the watch was,
during the three weeks we were cruising to the N.; in which interval, it
gave the longitude of the East Cape with a difference of twenty-eight
miles, I have marked the longitude of Saint Peter and Saint Paul, as given
by the time-keeper, notwithstanding it stopped a few days before we arrived
there; this I was enabled to do, from comparing the longitude it gave the
day before it stopped, with that given by Mr Bayley's watch, and allowing
for the error of the latter.

The use of so accurate a measure of time is sufficiently evident, from its
furnishing in itself the means of approximating to the longitude at sea, as
may be seen in the above table. But, besides this, we were enabled, by the
same means, to give a degree of accuracy to the lunar observations, which
they cannot otherwise pretend to; and, at the same time, by reducing a
number of those observations to one time, obtain results approaching still
nearer to the truth. In surveying coasts, and ascertaining the true
position of capes and head-lands, it reaches the utmost degree of practical
exactness. On the other hand, it is to be observed, that lunar
observations, in their turn, are absolutely necessary, in order to reap the
greatest possible advantages from the time-keeper; since, by ascertaining
the true longitude of places, they discover the error of its rate. The
original observations that were made in the course of this voyage, have
been published by order of the Board of Longitude, and to those I must
refer the reader, for his further information on this subject.

N.B. The observatories were placed on the west side of the village of Saint
Peter and Saint Paul.

Latitude deduced from meridian zenith
distances of the sun, and of five stars
to the S., and five to the N. of the
zenith 53 deg. 0' 38" N.
Longitude deduced from one hundred
and forty-six sets of lunar observations 158 43 16 E.
Longitudy by time-keeper, according to
its Greenwich rate 173 36 0
Longitude by time-keeper, according to
its rate found at Owhyhee 159 20 0
Variation of the compass, by azimuths
taken with three compasses, made by
Knigh