The scale fixed to the barometer is divided into inches, tenths, and half-tenths, so that each division on this scale is equal to 0.050 inch. The small movable scale or vernier attached to the instrument enables the observer to take more accurate readings; it is moved by a rack and pinion. Twenty-four spaces on the fixed scale correspond to twenty-five spaces on the vernier; hence each space on the fixed scale is larger than a space on the vernier by the twenty-fifth part of 0.050 inch, which is 0.002. Every long line on the vernier (marked 1, 2, 3, 4, and 5) thus corresponds to 0.010 inch. If the lower edge of the vernier coincides with a line on the fixed scale, and the upper edge with the twenty-fourth division of the latter higher up, the reading is at once supplied by the fixed scale as in A (Fig. 5), where it is 29.500 inches. If this coincidence does not take place, then read off the division on the fixed scale, above which the lower edge of the vernier stands. In B (Fig. 5) this is 29.750 inches. Next look along the vernier until one of its lines is found to coincide with a line on the fixed scale. In B this will be found to be the case with the second line above the figure “2.” The reading of the barometer is therefore:— On fixed scale 29.750 On vernier (12 × .002) .024 ------ Correct reading 29.774 Should two lines on the vernier be in equally near agreement with two on the fixed scale, then the intermediate value should be adopted. 5. Lower the mercury in the cistern by turning the screw at the bottom until the surface is well below the ivory point; this is done to prevent the collection of impurities on the surface about the point. The transport of barometers requires very great care in order to prevent the introduction of air into the tube or the fracture of the tube by the impact of the mercury against the top. To reduce the risk of these accidents, the barometer must be carried with the tube quite _full of mercury_, and in an inverted position, at least with the cistern end kept higher than the top of the tube. The flexible cistern of the Fortin type of barometer allows of it being screwed up tight so as to fill the tube and close the lower end of it. In case of breakage, the operation of fitting a new tube is not very difficult, but unless the tube has been carried out ready filled with mercury, this cannot well be attempted. In order to drive out the film of air adhering to the glass on the inside, it is necessary, after filling the tube, to raise its temperature to the boiling-point of mercury. No one should attempt either to fill or to change a barometer tube unless he has had practice in doing so under expert supervision beforehand.