seded by more simple forms. This is a matter, however, in
which great difficulties are found to the present day by Englishmen,
whose language presents no certain laws for rendering any given sound
into a fixed combination of letters.

Cook's language is unvarnished and plain, as a sailor's should be. His
incidents, though often related with circumstance, are without
exaggeration; indeed if any fault is to be found, it is that he takes
occurrences involving much labour and hardship as such matters of course,
that it is not easy for the reader, especially if he be a landsman, to
realise what they really entail.

Cook was assiduous in obtaining observations to ascertain the Variation
of the compass--i.e., the difference between the direction shown by the
magnetic needle and the true north. He is constantly puzzled by the
discrepancies in these observations made at short intervals. These arose
from the different positions of the ship's head, whereby the iron within
a certain distance of the compass is placed in different positions as
regards the needle working the compass card, the result being that the
needle is attracted from its correct direction in varying degree. This is
known as the Deviation of the compass. The cause of this, and of the laws
which govern it, were only discovered by Captain Flinders in 1805.
Happily for the navigators of those days, little iron entered into the
construction of ships, and the amount of the Deviation was not large,
though enough to cause continual disquiet and wonderment.

Cook's longitudes in this voyage are all given as west of Greenwich, not
divided into east and west, as is usual at this day. The latter system
again has only been adopted universally since his time.

Though Cook himself gives, at the beginning of the Journal, a note of the
method of reckoning days adopted, it may not be amiss to give further
explanation here.

It was the usual custom on board ships to keep what was known as Ship
time--i.e., the day began at noon BEFORE the civil reckoning, in which
the day commences at midnight. Thus, while January 1st, as ordinarily
reckoned, is from midnight to midnight, in ship time it began at noon on
December 31st and ended at noon January 1st, this period being called
January 1st. Hence the peculiarity all through the Journal of the p.m.
coming before the a.m. It results that any events recorded as occurring
in the p.m. of January 1st in the log, would, if translated into t