MEASURING DISTANCES WITH LINEN TAPE.

227

one point to another, providing he pays close attention to the gaits of his animal and notes the time consumed with each different gait.

In floating down a river a fairly satisfactory measure of the distance traveled can be obtained with currents of various velocities by timing floats over a measured distance in stretches of comparatively slow velocity, up to those in which the speediest rapids are encountered. The explorer may thus float down the stream, using a sketch-board or prismatic compass for direction, and by timing the boat from one course to another a fairly good survey may be made of the route traveled. Similar methods may be employed in ascertaining the time necessary to row or paddle a boat in still water or against streams of varying velocities, and by endeavoring to maintain a uniform rate in rowing or paddling it is possible by timing the courses to get a fair estimate of the distances.

In platting paced and timed surveys it will be found desirable to arrange a scale of pacing or timing. Thus, instead of transposing the number of paces into distance paced, a -scale should be prepared on which should be graduated paces instead of distances (Fig. 55). For example, for a man who paces a yard at each stride, if the scale of plotting is to be one mile to the inch, there will be 10 paces to every of an inch,

I

17.6

176

and 100 paces to every of an inch, so that by dividing an inch into 17.6 parts it will be equal to 100 paces, and lesser fractions can be interpolated. In the same manner, if a horse strides with the same foot a distance of 6 feet at each step, the inch may be divided into 88 parts, and each one of these will be equivalent to 10 strides. In similar manner a scale of time may be prepared, or, better still, in each case several scales for different strides or for different times. Thus, for a scale of one mile to one inch, 15 minutes' travel at the rate of 3 miles an hour will be represented by of an

inch, and the same time at the rate of 4 miles an hour will be represented by one inch.

97. Measuring Distances with Linen Tape.—Various methods have been adopted for measuring distances on secondary and tertiary traverses in dense woods where the underbrush is so thick as to preclude the use of the stadia, and where the work required is such as to render unnecessary the accuracy attained by the use of steel tape or chain with two chainmen (Art. 99). Under such conditions two plans have generally been adopted: one, running of traverse lines by the topographer, directions being obtained by prismatic compass or plane-table (Arts. 91 and 61), and distances by the aid of an assistant who drags a chain; the other, by directions in the same manner, but distances by pacing (Art. 95). As the topographer can see but a few yards ahead of him, he rarely sights to a fixed object, but on small-scale work finds it sufficient to sight in the direction in which the assistant has preceded him, dragging the chain.

A more satisfactory and far more accurate mode of measurement under such circumstances has been found to consist in measuring distances with a long linen tape. This is made of tailor's linen binding-tape obtained at dry-goods stores in spools of five hundred to one thousand feet in length, the best for this purpose being so finely woven and so smooth that it slips through the brush without catching, and is dragged ahead by one tapeman, the alignment of the tape giving the direction which the topographer is to sight for his azimuth. It is improved by immersion in boiling paraffin. The peculiarity of this apparatus consists in the fact that ordinarily the end of the tape will catch in brush and around trees, and tear and fray. To prevent this at narrow strip of celluloid, of the same dimensions as the tape, is sewed on its extreme end, the length of this celluloid appendage being from twelve to eighteen inches, and this causes it to slip between the bushes without becoming

tangled or twisted.

With such device numerous traverses

have been run in the Adirondack woods and plotted on a scale of 1 inches to a mile, with average closure errors of to inch in circuits of 5 to 15 miles periphery.

98. Odometer. The odometer is not a distance-measurer, but a revolution-counter; consequently a function of such distance-measuring is the circumference of a wheel, the number of revolutions of which are counted. This wheel may be one of a buggy or other light conveyance, preferably a front wheel, in order that the odometer which is attached to it may be clearly in view at all times; or the wheel may be attached to a light hand-barrow, so that it can be trundled along trails or other routes over which two- or four-wheeled conveyances could not be driven.

Distance-measuring by means of rolling a wheel over the surface and recording with the odometer the number of times the periphery of the wheel is applied to the surface may be done under the most favorable circumstances with nearly the accuracy of ordinary chain or stadia measuring. Such accuracy is not as great as that by the latter methods where they are carefully executed, but is sufficient for all purposes of distance-measuring where the results are to be plotted on a geographic map.

The errors inherent in this work are of four kinds:

1. Those due to the difficulty of reducing measures on an inclined surface to horizontal;

2. Failure of the odometer or counter to correctly record the number of the revolutions;

3. Slip or jolt in the wheel, due chiefly to striking stones, roots, and other obstacles; and

4. Errors resulting from failure to run the wheel in a direct line between two station points.

The first is perhaps the most serious, and as yet no satisfactory means have been devised whereby an instrument will record the changes in inclination passed over by an odometer

wheel. The second may be partially guarded against only by using the best form of odometer and by the traverseman counting the revolutions of the wheel at the same time as a check. The third is not susceptible to correction, and errors due to this cause will occur unless the surface of the road be of exceptional quality. The errors due to the fourth cause may be practically eliminated by great care in driving or trundling the wheel in a straight line where the road surface will permit. These and like errors inherent in odometric surveys may be so greatly reduced by careful work as to render them of small moment when the survey is to be platted on a geographic map, and where there is sufficient

FIG. 73.-DOUGLAS ODOMETER ATTACHED TO WHEEL.

control by triangulation, stadia, or other equally good method to which to adjust the odometer traverses and thus eliminate their errors. As a general rule the errors due to odometer

ODOMETER AND HAND-RECORDER.

231

measurement for this class of work are no greater than those introduced in the measurement of directions and due to the difficulty of plotting short road tangents to a small scale.

There are several forms of odometer, among the best of which is the Douglas odometer, so named after its inventor, Mr. E. M. Douglas of the U. S. Geological Survey. (Fig. 73.) This is firmly fixed to the axle of the wheel, and a cam is welded around the hub, the lift of the cam being of such height that as it strikes the lever of the odometer it raises this by just the amount sufficient to turn the cog-wheels within the instrument and move the index forward one division for each lift of the cam, corresponding to each revolution of the wheel. This odometer records revolutions directly, and a similar result may be obtained by the use of the ordinary printing-press counter, which may be suitably rigged on the axle of the wheel. The old form of pendulum. odometer is so unreliable as to be of practically no value at all for purposes of surveying.

Another form of odometer which has been found to be very satisfactory and accurate is the bell odometer (Fig. 74).

The record of this is in miles, tenths, and hundredths, instead of in number of revolutions. As a consequence it is manufactured for different diameters of wheels. Knowing, there