When using the 25-inch maps for determining azimuths it must be borne in mind that the side-lines are not, necessarily, due north and south. The Director-General of the Ordnance Survey, Southampton, will probably on application state the correction to be applied to the azimuths on this account, and this should be applied, of course, to each of the values obtained. If for any reason it is found necessary or desirable to make observations of the azimuths independently of the Ordnance Survey, full instructions as to the method of procedure may be found in an inexpensive instruction book[133] issued by the Board of Education. The instructions given on p. 49, § 3, are most generally applicable, and the form on p. 76 will be found very handy for recording and reducing the observations. In making observations of the angular elevation of the horizon a good theodolite is essential. Both verniers should be read, the mean taken, and then the telescope should be reversed in its Ys, reset, and both readings taken again. One setting and reading are of little use. The Ordnance Survey maps may also be employed _in a preliminary reconnaissance_ to obtain approximate values of the horizon elevations. This may be done by measuring the distances and contour-lines shown on the one-inch maps. This method, however, is only very roughly approximate owing to the fact that sharp but very local elevations close to the monuments may not appear on these maps and yet be of sufficient magnitude to cause large errors in the results. Where trees, houses, &c., top the horizon, they should, of course, be neglected and the elevation of the ground level, at that spot, taken. Should the top of the azimuth mark (stone, &c.) show above the actual horizon, its elevation should be recorded and not that of the horizon. Having measured the angular elevation of the horizon along the sight-line, it is necessary to convert this into actual zenith distance and to apply the refraction correction before the computations of declination can be made. The process of doing this and of calculating the declination will be gathered from the examples given below:-- _Data._ Monument:--E. circle Tregeseal, lat. 50° 8′ N. _i.e._ colat = 39° 52′. Alignment. Centre of circle to Longstone. Az. (from 25″ Ordnance Map). N. 66° 38′ E. Elevation of horizon (measured) 2° 10′. Reference to the May-Sun curve, given on p. 263, indicates that this is probably an alignment to the sunrise on May morning. Therefore, in determining the zenith distance, the correction for the sun’s semi-diameter (16′) must be taken into account, allowing that 2′ of the sun’s disc was above the horizon when the observation was made. _Zenith Distance_:-- Zenith distance of true horizon = 90° „ „ local „ = 90° - 2° 10′ = 87° 50′ Bessel’s tables show that refraction, at altitude 2° 10′, raises sun 17′. If 2′ of sun’s limb is above horizon, sun’s centre is 14′ below.