uny tribe, but they build their houses in the rocks;
and against sharpshooters in broken ground, guns can do little. Through
field-glasses it was possible to see the enemy dodging behind their
rocks, whenever the puffs of smoke from the guns told them that a shell
was on its way. Perhaps smokeless powder would have put a stop to this.
But in any case, the targets presented to the artillery were extremely
bad.

Where they really were of great service, was not so much in killing the
enemy, but in keeping them from occupying certain spurs and knolls. On
30th September, when the Royal West Kent and the 31st Punjaub Infantry
were retiring under considerable pressure, the British Mountain Battery
moved to within 700 yards of the enemy, and opened a rapid fire of
shrapnel on the high ground which commanded the line of retreat, killing
such of the tribesmen as were there, and absolutely forbidding the hill
to their companions.

In all rearguard actions among the mountains the employment of artillery
is imperative. Even two guns may materially assist the extrication of
the infantry from the peaks and crags of the hillside, and prevent by
timely shells the tribesmen from seizing each point as soon as it is
evacuated. But there is no reason why the artillery should be stinted,
and at least two batteries, if available, should accompany a brigade to
the attack.

Signalling by heliograph was throughout the operations of the greatest
value. I had always realised the advantages of a semi-permanent line
of signal stations along the communications to the telegraph, but I had
doubted the practicability of using such complicated arrangements in
action. In this torrid country, where the sun is always shining,
the heliograph is always useful. As soon as any hill was taken,
communication was established with the brigadier, and no difficulty
seemed to be met with, even while the attack was in progress, in sending
messages quickly and clearly. In a country intersected by frequent
ravines, over which a horse can move but slowly and painfully, it is the
surest, the quickest, and indeed the only means of intercommunication.
I am delighted to testify to these things, because I had formerly been a
scoffer.

I have touched on infantry and artillery, and, though a previous chapter
has been almost wholly devoted to the cavalry, I cannot resist the
desire to get back to the horses and the lances again. The question of
sword or lance as the cavalrym