econd time. It was again opened, and the
cause of its stopping appeared to be owing to the man having put some part
of the work badly together when he first opened it. Being again adjusted,
it was found to gain above a minute a-day; and, in the attempt to alter the
regulator and balance-spring, he broke the latter. He afterward made a new
spring; but the watch now went so irregularly, that we made no farther use
of it. The poor fellow was not less chagrined than we were at our bad
success; which, however, I am convinced, was more owing to the miserable
tools he was obliged to work with, and the stiffness his hands had
contracted from his ordinary occupation, than to his want of skill.

For the satisfaction of those who may wish to have a general view of its
rate of going, I have added the following table.

The first and second columns contain the dates when, and the names of the
places where its rate was observed. The third column contains the daily
error of its rate, so found from mean time. The fourth column has the
longitude of each place, according to the Greenwich rate; that is,
calculated on a supposition that the time-keeper had not varied its rate
from the time it left Greenwich. But as we had frequent opportunities of
ascertaining the variation of its daily error, or finding its new rate, the
fifth column has the longitude according to its last rate, calculated from
the true longitude of the place last departed from. The sixth is the true
longitude of the place deduced from astronomical observations made by
ourselves, and compared with those made by others, whenever such could be
obtained. The seventh column shews the difference between the fourth column
and the sixth in space; and the eighth the same difference in time. The
ninth shews the number of months and days in which the error, thus
determined, had been accumulating. The difference between the fifth and
sixth columns is found in the tenth, and shews the error of the time-
keeper, according to its rate last found, in space; and the eleventh the
same error in time. The twelfth contains the time elapsed in sailing from
the place where the rate was last taken, to the place whose longitude is
last determined. The thirteenth and fourteenth contain the state of the air
at the time of each observation.

As persons, unaccustomed to calculations of this sort, may find some
difficulty in comprehending the nature of the table, the two following
instances will more