798

THEORY AND PRACTICE OF ENGINEERING.

Book II.

Fig. 921.

eye-glass of the telescope A B, at any lineparallel to the horizon, as that of the line ab ,on the house C, so as to cover it exactly withthe thread of the telescope; then turning thetelescopes half round on the two pivots G, H, sothat the telescope A B, which was on the left,comes on the right hand ; then looking throughthe same telescope A B, if the thread stillcovers the horizontal line a b, it is a proof thatthe telescope A B is well centred, and that itsvisual rays are parallel to the two pivots G andH : should it either cut above or below, thescrew at E must then be turned until the visualray cuts half way between e and/; then turn thetelescope round on the pivots, so that the telescope through which the observation hasbeen made may pass from right to left; on looking through it, if the thread cuts theline ef it is a proof it is well centred. The same rule must be observed with the othertelescope, and both being properly centred, it is only then necessary to have the two visualrays parallel with the horizon, and this is done by putting the weight into equilibrium :after this is done, point the telescope towards the object, as the house C, and if in lookingthrough the eye-glass of the telescope A B, its thread cuts the object in any point, as G.and having turned the whole level to make the eye-glass D of the telescope C D come tothe same point, it is then a proof the visual rays are parallel with the horizon. 'Hie weightN, by being advanced or pushed back, changes at once the equilibrium of the level, and bythe whole it may be truly and correctly adjusted.

Level lines and points are all equidistant from the centreof the earth ; the points B, K, L, and C, are equidistantfrom the centre of the globe A. There are two de-scriptions of levels, one called the true, the other the ap-parent : the true level is that which follows the circum-ference ; the apparent is that which is drawn perpen-dicular to the radius of the circle: B G I F, which is sodrawn at the end of A B, is the apparent level.

The term levelling implies the idea of bringing anything to a level or flat; but, as engineers use it, it simplymeans the process by which the quantity of deviation froma true level is ascertained, the object being to find out ifthe surface be raised or sunk, or what is the slope or in-clination. Level lines or level planes are supposed to beperfectly flat, or parallel with the horizon where they exist,and a sheet of water at rest may be supposedto have its surface level, for if any part ofthe plane on which it rested was lower thananother, it would run in that direction.

It therefore follows that some allowanceshould be made in levelling through very longdistances, as from E to D, D to E, and B to A,and for this very accurate tables have been com-pleted for all distances within the range of anobservation.

We have also to make some allowances forthe refraction, which varies from £ to ^ of theangle subtended by the horizontal distance ofobjects. In the ordinary state of the atmo-sphere, the refraction is about a fourteenth ofthe horizontal angle, and the radius of the cur-vature of the ray seven times that of the earth.

The effect of refraction may be allowed for bycomputing the correction for curvature, and Tig. 924-

then taking one-seventh for the quantity by which the object is rendered higher by therefraction than it ought to be.

Numerous tables have been drawn up to enable the engineer to correct the curvature andrefraction for distances in chains, feet, or miles ; the corrections for refraction are takenusually from one-seventh up to a twelfth of the apparent above the true level, which isaffected by the state of the atmosphere.

Suppose a spring to issue from the earth at G, the water would not flow along the lineH F, which is the apparent level, but would flow towards K and L, in the circle N, and

Fig. 922.

Fig. 923.