OoMRASS SURVEY. 79 
both cases it should be stated on the map whether the meridian is true 
or magnetic. 
Each station where bearings are taken must be plotted in a similar 
manner to Camp 2, that is, by bearing from the last station, and by 
estimated distance. Having by means of the first two stations fixed hills 
off the line of march, bearings of these will assist to obtain the position 
of the third, and so on. When no object can be seen to march for, the 
direction must be obtained by compass bearing of the line of march 
obtained from time to time. 
The aneroid readings, and the boiling-point, furnish us with the means 
of ascertaining approximately the difference in height of two stations, 
which may be computed by the tables (see pp. 210 to 213, 217, 218), or, 
where the height is not considerable, by a simple arithmetical process as 
follows 
Take the sum and difference of the aneroid readings, at the upper and 
lower station, get the mean of the temperature in the shade at the two 
stations. Then, sum of readings: difference of readings: : 55,000: the 
difference in height. Increase the result thus found by of itself for 
every degree that the mean temperature in the shade at the two stations 
exceeds 55°; subtract the like amount if it is below 55°. The aneroid 
readings, in the example, computed by the tables and this formula, will 
show a fairly close agreement. 
Approximate 
Method. 
By Tables. 
Feet. 
Feet. 
A, above Camp 1 .. 
.. 1603*8 
1st Camp above 2 nd Camp .. 
.. 308*8 
Foot of Range above 2 nd Camp .. 
.. 477 ' 2 “ 
•* 475*9 
Height of Range E. 
.. 1145*0 
„ by Boiling point 
.. 
.. 1155*3 
E above Lake .. .. . 
•• 956*5 
Eor plotting the work in the field, a scale of one inch to the geographical 
anile will exhibit all the main features of a country traversed in a day's 
journey* Special plans must be drawn on a scale suited to the area they 
are intended to represent; but whatever scale is chosen for the field work, it 
should be large enough to admit of considerable reduction in the fair plan, 
as by this process all errors are diminished. The projection of maps is pur¬ 
posely omitted here, as it is dealt with separately (see p. 58 et seg .); it will, 
