196 
MINUTES OF PROCEEDINGS OF 
Cutoff the 100, or if there is 200 in the answer, cut off the 200. 
Final result 57. 
Therefore the instruments read 57 too high or 43 too low. 
When the instruments are in order the second result should he 100 or 
200 . 
In this last test it may be as well to point out that there is 
a slight error if the point in front he near, that is, if the point he distant 
2,000 yards, instruments which apparently add up to 100, will, in reality, 
add up to 100J ; if the point be distant 1,000 yards, then, while they 
apparently add up to 100, they will really be set at 100J. 
The sun test is mathematically true. If the instruments are set 
exactly at 100, we are, in measuring a distance of 2,000 yards, liable to a 
plus error equal to a division of J, at 1000 to a plus error of J. 
If the bar is in the centre of the divided plate, when the distance of 
an object is being taken, or within one quarter of the whole divided 
plate from the centre point, the obliquity of the triangle practically 
introduces no error; the bar need only he looked at to ascertain this. 
As a rough test, the range of the same object may be taken with two 
widely different bases, as 30 and 60 yards ; if the answers agree, the 
instruments must be in order. 
The result of each careful test should be written inside the lid of the 
box, with the date and name of the worker; this will provide a history of 
the instrument. 
Instruments testing between 974 and 100 are in good condition, but 
they should never be allowed to test too high. 
Manner of inscribing Tests on Box. 
Tested 10th Jan., 1870, 97J. Lieut. A. 
„ by Sun, 7th Nov. 1870, 97f. Captain B. 
„ 7th July, 1871, 97J. Captain B. 
To Correct a Pair of Instruments. 
Let us suppose that we have, by the above tests, discovered that a pair 
of instruments read 5 too much. 
How is this 5 to be corrected ? If in a great hurry, the gun mark of 
the roller may be scratched out, and a new one made 5 below (at 95). 
The tape must always be used if this has been done.* It is a bad hut a 
ready plan; the answers are perfectly accurate. 
The second system of correction takes more time, but is infinitely 
better, not in accuracy, but in simplicity. The instruments read 5 too 
high, we want to make them read 5 lower. 
Choose two distant objects tolerably square with one another. 
Lay the telescopes of one instrument on both. Write down the gun 
number, say, 47 J. 
If we could make the instruments now read 42J, they would be 
correct. 
To effect this— 
Take off the case. 
That is, observations must not be taken on the muzzle. 
