Should the elevation of the place where the barometer is to be used be considerably above the sea level, it will be well--after suspending it from the stand--to unscrew the cistern several turns, _holding the barometer in an oblique position_, as at great heights the mercury will fall considerably quicker than the cistern can be unscrewed, thereby filling it to overflowing; but by partly unscrewing the cistern first, room is given for the reception of a fall of mercury to the extent of several inches. The cistern must not be unscrewed when the _Instrument is_ INVERTED _more than_ two turns of the screw, otherwise the mercury will flow out through the groove. It is found safer when travelling to carry the barometer in a horizontal position, or with its cistern end uppermost. _To clean the Barometer._--Should at any time the mercury in the cistern become oxidised, and reading from its surface be difficult, it can be readily cleaned by removing the cistern and its contained mercury from the barometer frame by unscrewing it _when in a horizontal position_; this precaution is necessary that the mercury in the tube may not escape, and thereby allow air to enter. The cistern must then be emptied, and with a dry clean leather, or silk handkerchief, well cleaned. The operation of cleaning being performed, return the cistern to the frame, and screw it until the face is brought up against the under side of the shoulder, still keeping the instrument _horizontal_. The cistern is now ready for re-filling, to do which stand the barometer on end _head downwards_, and remove the small screw at bottom; through the aperture thus opened, pour in mercury, passing it through a paper funnel with a very small aperture. It is well to pass the mercury through a very small funnel two or three times before returning it to the barometer cistern, as by this process all particles of dust or oxide adhere to the paper, and are effectually removed. Should any small quantity of the mercury be lost during the operation of cleaning, it is of no importance so long as sufficient remains to allow of adjustment to the zero point. This latter constitutes one of the great advantages of this new instrument over the ordinary barometer; for, in the majority of cases, after an instrument has been compared carefully with a standard, should mercury be lost, there is no means of correcting the error unless a standard barometer be at hand; the new barometer is, in this respect, independent, a little mercury more or less being unimportant. =41. Short Tube Barometer.=--This is simply a tube shorter, as may be required, than that necessary to show the atmospheric pressure at the sea level. It is convenient for balloon purposes, and for use at mountain stations, being of course a special construction.