110 
HINTS TO TRAVELLERS. 
for short distances, and I could generally guess a bearing to within 
2° or 3° of the truth. This error in short distances, when the route is 
not plotted to any large scale, is of no importance. To find the distance, 
I noted the time taken in traversing each by a watch reading seconds, 
occasionally pacing one hundred yards to find the rate of going, all halts 
or checks, of course, being noted also. 
“By this method, frequent stoppages of the whole line in a narrow 
path, from which it was impossible to step aside to take compass 
readings, were avoided. The compass is often affected by the proximity 
of arms and accoutrements, and this difficulty is also overcome. The 
changes in the direction of a path through jungle, or on a hillside, where 
there is no made road, are very frequent; and observations of shadows 
enable one to determine, without observing the compass, whether the 
direction of the path really changes, or only alters for a few yards, 
resuming the old course again. Accurate measurements by pacing are 
only obtainable by keeping up a continuous steady walk, which it is 
impossible to do with the frequent checks, or spasmodic accelerations of 
pace on a line of march; but I found by repeated trials that the rate of 
a column does not vary nearly so greatly as the pace of any one individual 
in it. Considerable practice is necessary to acquire accuracy in steep 
ground, but in tolerably easy country I found I could easily obtain it. 
Fortunately for this method, all countries are not so sunless as England. 
On one occasion, 1 made a route survey in this way for about forty miles 
of hill and dale, with only one check ray to a known point; and when it 
was transferred to an accurate survey, which was afterwards made of the 
country traversed by it, its last station was found to be hardly out at all 
in latitude, and not half a mile in longitude. In the cold weather of 
1876-77,1 had to survey some rapid shallow streams running through 
dense jungle, and whenever we were going with the stream in our dug- 
outs (7.e., native boats, each hollowed out of a single tree), I found the 
best plan of surveying was with a prismatic compass, suspended in 
gimbals mounted on a small tripod-stand set up in front of my seat in 
the boat. I measured certain distances along the bank, and carefully 
noted the time my boat took to pass them, carried down by the current 
only. The compass gave the bearing throughout the length of the reach, 
and the watch gave the distance, and I found quite sufficiently accurate 
results were obtained. In actual measurements of shallow streams, when 
