When the barometer is placed in the ship, its position should be as near midships as possible, out of the reach of sunshine, but in a good light for reading, and in a situation in which it will be but little liable to sudden gusts of wind and changes of temperature. Great care should be taken to ascertain the exact height of its cistern above the water-line, and in order to facilitate night observations every possible arrangement should be made for placing behind it a light screened by white paper.
_Observations._--The first thing to be done is the reading off and recording the temperature indicated by the thermometer that in this construction of instrument dips into the mercury in the cistern. Sir John Herschel has suggested that "the bulb of the thermometer should be so situated as to afford the best chance of its indicating the exact mean of the whole barometric column, that is to say, fifteen inches above the cistern enclosed within the case of the barometer, nearly in contact with its tube, and with a stem so long as to be read off at the upper level."
Previous to making an observation with the barometer the instrument should be slightly tapped to free the mercury from any adhesion to the glass; any violent oscillation should, however, be carefully avoided. The vernier should then be adjusted to the upper surface of the mercury in the tube; for this purpose its back and front edges should be made to coincide, that is, the eye should be placed in exactly the same plane which passes through the edges; they should then be brought carefully down until they form a tangent with the curve produced by the convex surface of the mercury and the light is _just_ excluded from between them and the point of contact. It is desirable in making this adjustment that the eye should be assisted by a magnifying-glass. The reading of the scale should then be taken and entered in the column appropriated to it in the proper form. If the instrument have no tubular or double-edged index, the eye should be placed carefully at the level of the upper surface of the mercury and the index of the vernier brought gently down to the same level so as apparently just to touch the surface, great care being taken that the eye index and surface of the mercury are all in the same plane.