Rangefinders.—Some rangefinders are stereoscopic in principle and resemble, in that respect, the stereoscopic plotting machines with their "floating mark." Coincidence methods have the great advantage over stereoscopic that almost everyone can use them. Moreover for this particular purpose a general plastic image is not required. Errors involved in the use of the rangefinder, for any given base and magnification, are proportional to the square of the distance and an error of 4 may be taken as normal at a range of 250 yards with a 31.5 inch (8 cm.) magnification 8 model. Models reach roo feet in base length and 28 in magnification. With a metre base and magnification 20 the fractional error is less than 60o at 500 yards. An increasing survey use is being made of this rangefinder. In Canada, Egypt and elsewhere surveyors are finding its advantage over the tachymeter in the independence of rodmen and the increased speed. In Canada it has been, and is, used in the control of river and lake traverses on which the surveys, nowa days mainly air photo-surveys, of the northern territories are based. Obviously the rangefinder would prove useful in medium scale plane tabling for the measurement round each resected point.
Geodetic levelling would not be available, normally, as a backbone or control for a topographical survey. For this latter purpose secondary levelling with errors not exceeding about 0.03 foot per mile would suffice. These secondary chains, spaced fifty to a hundred miles apart in a grid or net, are generally confined to railways or roads. The intervals would then be filled up with subsidiary orders of levels reaching eventually such errors as 0.1 foot per mile over short lengths. Levels of a topographical class are often run with the tachymeter in conjunction with plane tabling.
In topographical surveys of new countries it is usual to base contouring on a control of heights determined trigonometrically (by vertical angles with a theodolite). The operations of triangu lation and determination of heights are thus combined. Truly simultaneous observation between two points will tend to elimi nate refraction. In all other cases there is bound to be uncer tainty. Observation of vertical angles should be confined to the times of minimum refraction (early in the afternoon), and heights should be carried forward through the shortest sides—since the correction for refraction varies as the square of the distance.
Trigonometrical heights show errors of the order of two or three feet in ioo miles.
In plane-table surveys minor heights are obtained with the Indian pattern clinometer, a small instrument which reads slopes, vertical angles, and natural tangents of the vertical angles. The clinometer (nine inches long) has two vanes. In the rear vane is a sight hole which can be levelled in relation to the zero point of the front vane. If confined to a radius of two miles this clinometer gives excellent results, but at longer ranges errors rapidly increase.
The increasing use of the aneroid barometer is due to the con venience of measuring height directly instead of deducing it from distance and slope. A barometer measures the pressure of the air however and before it can be used to measure either absolute or relative heights corrections must be applied to eliminate the effect of (a) Weather, diurnal, and area variations, (b) Tempera ture, (c) Index or other instrumental errors.
The first correction is usually applied by comparing the record of a barometer retained in camp with that of the field instrument. The second can be applied directly if the shade temperature is measured at the time of each reading, but thermometer readings must also be taken for the stationary or camp instrument. Under particularly stable conditions it is possible to record the movement due to average diurnal variations and to dispense with the camp instrument, the recording of which is often difficult to arrange. In these cases the record graphs should be either corrected for variation in temperature or amplified by a statement of the mean temperature at definite time intervals. There is much to be done still in designing survey barometers with an eye to reducing instru mental errors such as lag, friction of pivots, etc.
The hypsometer (or boiling point apparatus) is of more value to the explorer than the surveyor.
In topographical surveying contours are not generally followed out in detail, but are sketched in upon the control. Strict accuracy in contouring is not perhaps a very important feature of a topo graphical map. Topographical scales will not show, in any case, enough information for engineering enterprise, whereas a rela tively good and reliable picture of the accidents of the ground is of the first order of importance in the development of a new country. The sketching in of the contours is helped by measuring gradients, by noting all points in the observer's level and by spac ing contours properly on rivers and streams. There is little dif ficulty in attaining sensible accuracy in the position of contours .in hilly country, but errors of four or five feet in altitude are normal and may be appreciable in flat country.