or distance within $1\frac{1}{2}'$ of position from
a correct result; in finding the ship's
position by Sumner’s method a margin
of $2\frac{1}{2}'$ will be allowed.

(b.) In such problems as the “amplitude”
and “altazimuth,” where the bearing,
deviations, &c., only are required, a
margin of 3' or 4' from a correct result
will be sufficiently accurate.

(c.) Candidates for ordinary certificates are not
required to correct for second differences
in taking out the quantities from the
“Nautical Almanac”; and even candidates
for extra certificates are only required
to show that they are acquainted with the
method of second differences by correcting
the elements for same in the lunar problem.

(d.) In solving the time-azimuth problems an
answer not exceeding half a degree from
the exact result will be sufficiently near.
But in all cases the actual latitude,
declination, and time used, together with
the exact bearing from the north or south
as given in the tables, must be clearly
shown by the candidate on his papers.

(e.) In computing the time at which a given star
will be on the observer's meridian, and
the name of the stars near the meridian
(sections $b$ and $e$ of par. 33), an approximation
only is required, and it will be
sufficiently precise if the candidate works
throughout with the nearest minute of time.
In computing the approximate meridian altitude
of a star (section $g$), working throughout
with the nearest minute of arc will also be sufficiently close.

(f.) In interpolating for the correct deviation
to be applied in solving the chart questions,
it will usually be sufficient if the candidate
works throughout with the nearest degree
of deviation taken from the deviation-card;
and even in cases where the deviations may
vary but little, the nearest half-degree used
throughout will be sufficiently precise.
It is not necessary that the candidate should
waste his time in solving the course to odd minutes,
as is sometimes done.

(g.) In calculating the correction to apply to
soundings the candidate is not required to
work to the exact inch, as is sometimes done.
It will be sufficient if he brings his answer
within half a foot or so of a precise result.

(h.) In the lunar problem a difference of 3' in
the longitude from the correct answer as
deduced by a rigorous method by spherical
trigonometry will be allowed.

It must be clearly understood, in reading the
foregoing instructions as to the precision required,
that they only apply when the work of the
candidate is correct in principle.

183. A candidate will not be allowed to undergo
     examination twice in the same week, unless, under
     very special and urgent circumstances, the Marine
     Department sees fit to relax this rule. In that case
     a different set of problems should be given to
     the candidate.

APPENDICES.

APPENDIX A.

THE SIGHT-TESTS.

1. FORM-VISION TEST.

(1.) THE test for form-vision is the first test
which the candidate is required to undergo, and until
he has passed this test he cannot be allowed to proceed
further with the examination. (See par. 11.)

(2.) The test to be used are Snellen’s letter test
for candidates who can read, and the “dot” tests for
those who cannot read.
The sets of tests which have been supplied to the
Examiners consist respectively of eight sheets of
Snellen’s letters and two sheets of dots.

(3.) The chief object of the tests for form-vision
is to show whether the candidate possesses eyesight
of sufficient strength and range, or, in other words,
they are means of discovering whether the candidate
has good or bad sight.
They also afford a means of detecting whether
a candidate is suffering from that form of colour-blindness
which is caused by the excessive use of tobacco,
and by illness or similar affections. All
candidates who are suffering from colour-blindness
arising from causes of that nature will be found
to be incapable of passing the tests for form-vision.

(4.) Candidates may use both eyes or either eye
when being tested, but they must not be allowed
to use spectacles or glasses of any kind.
The set of tests is to be hung on the wall, in a good
light, at a height of about 5 ft. or 6 ft. from the
ground.

(5.) If the candidate can read, he should be placed
at a distance of 16 ft. from the letter-test sheets,
and exactly opposite them. The distance should
never be varied, and the candidate should not be
allowed to approach within 16 ft. of the test-sheets.
One of the eight sheets of letters should then be
exposed, and the candidate should be asked to read
the letters, beginning at the top and going downwards.
If he can read correctly three of the five letters
in the fifth line from the top (third from the bottom),
or four of the letters in either of the two lines below
he may be considered to have passed the test. If
he cannot do so he should be treated as having failed.
Candidates should be tested with at least two of
the sheets of letters, and the same two sheets should
not be used for all candidates.

(6.) The number printed over each row of letters
only shows the distance in feet at which a person
possessing ordinary or normal sight should be able
to read the letters in the row, and does not apply
to the distance at which a candidate under examination
should be tested.

(7.) If the candidate cannot read he must be tested
with the sheets of dots. For this test he is to be
placed at a distance of precisely 8 ft. from the
test-sheets, and exactly opposite them. One of
the sheets of dots is then to be exposed, and the
candidate should be asked to name the number of
dots in one or two of the lines or groups. Lines
and groups of dots can be formed by holding a piece
of white paper over part of the sheet, but care must
be taken that when this is being done the candidate’s
view is not obstructed, or the light on the test-sheet
in any way obscured.