At noon next day, in the cantonments at Jelapahar, an officer saw a strange sight--a field-hospital doolie with the red cross, and twelve _kahars_, Lucknow men, whose plaintive chant must have recalled old days on the North-West frontier. Behind on a mule rode a British orderly of the King's Own Scottish Borderers, bearded and weather-stained, and without a trace of the spick-and-spanness of cantonments. I saw the officer's face lighten; he became visibly excited; he could not restrain himself--he swung round, rode after my orderly, and began to question him without shame. Here was civilization longing for the wilderness, and over there, beyond the mist, under that snow-clad peak, were men in the wilderness longing for civilization.

A cloud swept down and obscured the Jelap, as if the chapter were closed. But it is not. That implacable barrier must be crossed again, and then, when we have won the most secret places of the earth, we may cry with Burton and his Arabs, 'Voyaging is victory!'

CHAPTER VIII

THE ADVANCE OF THE MISSION OPPOSED

The intention of the Tibetans at the Hot Springs has not been made clear. They say that their orders were to oppose our advance, but to avoid a battle, just as our orders were to take away their arms, if possible, without firing a shot. The muddle that ensued lends itself to several interpretations, and the Tibetans ascribe their loss to British treachery. They say that we ordered them to destroy the fuses of their matchlocks, and then fired on them. This story was taken to Lhasa, with the result that the new levies from the capital were not deterred by the terrible punishment inflicted on their comrades. Orders were given to oppose us on the road to Gyantse, and an armed force, which included many of the fugitives from Guru, gathered about Kangma.

The peace delegates always averred that we fired the first shot at Guru.

But even if we give the Tibetans the benefit of the doubt, and admit that the action grew out of the natural excitement of two forces struggling for arms, both of whom were originally anxious to avoid a conflict, there is still no doubt that the responsibility of continuing the hostilities lies with the Tibetans.

On the morning of April 7 ten scouts of the 2nd Mounted Infantry, under Captain Peterson, found the Tibetans occupying the village of Samando, seventeen miles beyond Kalatso. As our men had orders not to fire or provoke an attack, they sent a messenger up to the walls to ask one of the Tibetans to come out and parley. They said they would send for a man, and invited us to come nearer. When we had ridden up to within a hundred yards of the village, they opened a heavy fire on us with their matchlocks. Our scouts spread out, rode back a few hundred yards, and took cover behind stones. Not a man or pony was. .h.i.t. Before retiring, the mounted infantry fired a few volleys at the Tibetans who were lining the roofs of two large houses and a wall that connected them, their heads only appearing above the low turf parapets. Twice the Tibetans sent off a mounted man for reinforcements, but our shooting was so good that each time the horse returned riderless. The next morning we found the village unoccupied, and discovered six dead left on the roofs, most of whom were wounded about the chest. Our bullets had penetrated the two feet of turf and killed the man behind. Putting aside the question of Guru, the Samando affair was the first overt act of hostility directed against the mission.

After Samando there was no longer any doubt that the Tibetans intended to oppose our advance. On the 8th the mounted infantry discovered a wall built across the valley and up the hills just this side of Kangma, which they reported as occupied by about 1,000 men. As it was too late to attack that night, we formed camp. The next morning we found the wall evacuated, and the villagers reported that the Tibetans had retired to the gorge below. This habit of building formidable barriers across a valley, stretching from crest to crest of the flanking hills, is a well-known trait of Tibetan warfare. The wall is often built in the night and abandoned the next morning. One would imagine that, after toiling all night to make a strong position, the Tibetans would hold their wall if they intended to make a stand anywhere. But they do not grudge the labour. Wall-building is an instinct with them. When a Tibetan sees two stones by the roadside, he cannot resist placing one on the top of the other. So wherever one goes the whole countryside is studded with these monuments of wasted labour, erected to propitiate the genii of the place, or from mere force of habit to while away an idle hour. During the campaign of 1888 it was this practice of strengthening and abandoning positions more than anything else which gained the Tibetans the reputation of cowardice, which they have since shown to be totally undeserved.

On April 8, owing to the delay in reconnoitring the wall, we made only about eight miles, and camped. The next morning we had marched about two miles, when we found the high ridge on the left flank occupied by the enemy, and the mounted infantry reported them in the gorge beyond.

Two companies of the 8th Gurkhas under Major Row were sent up to the hill on the left to turn the enemy's right flank, and the mountain battery (No. 7) came into action on the right at over 3,000 yards. The enemy kept up a continuous but ineffectual fire from the ridge, none of their jingal bullets falling anywhere near us. The Gurkhas had a very difficult climb. The hill was quite 2,000 feet above the valley; the lower and a good deal of the other slopes were of coa.r.s.e sand mixed with shale, and the rest nothing but slippery rock. The summit of the hill was approached by a number of step-like shale terraces covered with snow. When only a short way up, a snowstorm came on and obscured the Gurkhas from view. The cold was intense, and the troops in the valley began to collect the spa.r.s.e brushwood, and made fires to keep themselves warm.

On account of the nature of the hillside and the high alt.i.tude, the progress of the Gurkhas was very slow, and it took them nearly three hours to reach the ridge held by the enemy. When about two-thirds of the way up, they came under fire from the ridge, but all the shots went high. The jingals carried well over them at about 1,200 yards. The enemy also sent a detachment to meet them on the top, but these did not fire long, and retired as the Gurkhas advanced. When the 8th reached the summit, the Tibetans were in full flight down the opposite slope, which was also snow-covered. Thirty were shot down in the rout, and fifty-four who were hiding in the caves were made prisoners.

In the meanwhile the battery had been making very good practice at 3,000 yards. Seven men were found dead on the summit, and four wounded, evidently by their fire.

But to return to the main action in the gorge. The Tibetans held a very strong position among some loose boulders on the right, two miles beyond the gully which the Gurkhas had ascended to make their flank attack. The rocks extended from the bluff cliff to the path which skirted the stream. No one could ask for better cover; it was most difficult to distinguish the drab-coated Tibetans who lay concealed there. To attack this strong position General Macdonald sent Captain Bethune with one company of the 32nd Pioneers, placing Lieutenant Cook with his Maxim on a mound at 500 yards to cover Bethune's advance. Bethune led a frontal attack. The Tibetans fired wildly until the Sikhs were within eighty yards, and then fled up the valley. Not a single man of the 32nd was. .h.i.t during the attack, though one sepoy was wounded in the pursuit by a bullet in the hand from a man who lay concealed behind a rock within a few yards of him. While the 32nd were dislodging the Tibetans from the path and the rocks above it, the mounted infantry galloped through them to reconnoitre ahead and cut off the fugitives in the valley. They also came through the enemy's fire at very close quarters without a casualty.

On emerging from the gorge the mounted infantry discovered that the ridge the Tibetans had held was shaped like the letter S, so that by doubling back along an almost parallel valley they were able to intercept the enemy whom the Gurkhas had driven down the cliffs. The unfortunate Tibetans were now hemmed in between two fires, and hardly a man of them escaped.

The Tibetan casualties, as returned at the time, were much exaggerated.

The killed amounted to 100, and, on the principle that the proportion of wounded must be at least two to one, it was estimated that their losses were 300. But, as a matter of fact, the wounded could not have numbered more than two dozen.

The prisoners taken by the Gurkhas on the top of the ridge turned out to be impressed peasants, who had been compelled to fight us by the Lamas.

They were not soldiers by inclination or instinct, and I believe their greatest fear was that they might be released and driven on to fight us again.

The action at the Red Idol Gorge may be regarded as the end of the first phase of the Tibetan opposition. We reached Gyantse on April 11, and the fort was surrendered without resistance. Nothing had occurred on the march up to disturb our estimate of the enemy. Since the campaign of 1888 no one had given the Tibetans any credit for martial instincts, and until the Karo la action and the attack on Gyantse they certainly displayed none. It would be hard to exaggerate the strategical difficulties of the country through which we had to pa.s.s. The progress of the mission and its escort under similar conditions would have been impossible on the North-West frontier or in any country inhabited by a people with the rudiments of sense or spirit. The difficulties of transport were so great that the escort had to be cut down to the finest possible figure. There were barely enough men for pickets, and many of the ordinary precautions of field manoeuvres were out of the question.

But the Tibetan failed to realize his opportunities. He avoided the narrow forest-clad ravines of Sikkim and Chumbi, and made his first stand on the open plateau at Guru. Fortunately for us, he never learnt what transport means to a civilized army. A bag of barley-meal, some weighty degchies, and a ma.s.sive copper teapot slung over the saddle are all he needs; evening may produce a sheep or a yak. His movements are not hampered by supplies. If the importance of the transport question had ever entered his head, he would have avoided the Tuna camp, with its Maxims and mounted infantry, and made a dash upon the line of communications. A band of hardy mountaineers in their own country might very easily surprise and annihilate an ill-guarded convoy in a narrow valley thickly forested and flanked by steep hills. To furtively cut an artery in your enemy's arm and let out the blood is just as effective as to knock him on the head from in front. But in this first phase of the operations the Tibetans showed no strategy; they were badly led, badly armed, and apparently devoid of all soldier-like qualities. Only on one or two occasions they displayed a desperate and fatal courage, and this new aspect of their character was the first indication that we might have to revise the views we had formed sixteen years ago of an enemy who has seemed to us since a unique exception to the rule that a hardy mountain people are never deficient in courage and the instinct of self-defence.

The most extraordinary aspect of the fighting up to our arrival at Gyantse was that we had only one casualty from a gunshot wound--the Sikh who was shot in the hand at the Dzama Tang affair by a Tibetan whose jezail was almost touching him. Yet at the Hot Springs the Tibetans fired off their matchlocks and rifles into the thick of us, and at Guru an hour afterwards the Gurkhas walked right up to a house held by the enemy, under heavy fire, and took it without a casualty. The mounted infantry were exposed to a volley at Samando at 100 yards, and again in the Red Idol Gorge they rode through the enemy's fire at an even shorter range. In the same action the 32nd made a frontal attack on a strong position which was held until they were within eighty yards, and not a man was. .h.i.t. No wonder we had a contempt for the Tibetan arms.

Their matchlocks, weapons of the rudest description, must have been as dangerous to their own marksmen as to the enemy; their artillery fire, to judge by our one experience of it at Dzama Tang, was harmless and erratic; and their modern Lhasa-made rifles had not left a mark on our men. The Tibetans' only chance seemed to be a rush at close quarters, but they had not proved themselves competent swordsmen. My own individual case was sufficient to show that they were bunglers. Besides the twelve wounds I received at the Hot Springs, I found seven sword-cuts on my poshteen, none of which were driven home. During the whole campaign we had only one death from sword-wounds.

Arrived at Gyantse, we settled down with some sense of security. A bazaar was held outside the camp. The people seemed friendly, and brought in large quant.i.ties of supplies. Colonel Younghusband, in a despatch to the Foreign Office, reported that with the surrender of Gyantse Fort on April 12 resistance in that part of Tibet was ended. A letter was received from the Amban stating that he would certainly reach Gyantse within the next three weeks, and that competent and trustworthy Tibetan representatives would accompany him. The Lhasa officials, it was said, were in a state of panic, and had begged the Amban to visit the British camp and effect a settlement.

On April 20 General Macdonald's staff, with the 10-pounder guns, three companies of the 23rd Pioneers, and one and a half companies of the 8th Gurkhas, returned to Chumbi to relieve the strain on the transport and strengthen the line of communications. Gyantse Jong was evacuated, and we occupied a position in a group of houses, as we thought, well out of range of fire from the fort.

Everything was quiet until the end of April, when we heard that the Tibetans were occupying a wall in some strength near the Karo la, forty-two miles from Gyantse, on the road to Lhasa. Colonel Brander, of the 32nd Pioneers, who was left in command at Gyantse, sent a small party of mounted infantry and pioneers to reconnoitre the position. They discovered 2,000 of the enemy behind a strong loopholed wall stretching across the valley, a distance of nearly 600 yards. As the party explored the ravine they had a narrow escape from a b.o.o.by-trap, a formidable device of Tibetan warfare, which was only employed against our troops on this occasion. An artificial avalanche of rocks and stones is so cunningly contrived that the removal of one stone sends the whole engine of destruction thundering down the hillside. Luckily, the Tibetans did not wait for our main body, but loosed the machine on an advance guard of mounted infantry, who were in extended order and able to take shelter behind rocks.

On the return of the reconnaissance Colonel Brander decided to attack, as he considered the gathering threatened the safety of the mission. The Karo Pa.s.s is an important strategical position, lying as it does at the junction of the two roads to India, one of which leads to Kangma, the other to Gyantse. A strong force holding the pa.s.s might at any moment pour troops down the valley to Kangma, cut us off in the rear, and destroy our line of communications. When Colonel Brander led his small force to take the pa.s.s, it was not with the object of clearing the road to Lhasa. The measure was purely defensive: the action was undertaken to keep the road open for convoys and reinforcements, and to protect isolated posts on the line. The force with the mission was still an 'escort,' and so far its operations had been confined to dispersing the armed levies that blocked the road.

On May 3 Colonel Brander left Gyantse with his column of 400 rifles, comprising three companies of the 32nd Pioneers, under Captains Bethune and Cullen and Lieutenant Hodgson; one company of the 8th, under Major Row and Lieutenant Coleridge, with two 7-pounder guns; the Maxim detachment of the Norfolks, under Lieutenant Hadow; and forty-five of the 1st Mounted Infantry, under Captain Ottley. On the first day the column marched eighteen miles, and halted at Gobshi. On the second day they reached Ralung, eleven miles further, and on the third marched up the pa.s.s and encamped on an open spot about two miles from where the Tibetans had built their wall. A reconnaissance that afternoon estimated the enemy at 2,000, and they were holding the strongest position on the road to Lhasa. They had built a wall the whole length of a narrow spur and up the hill on the other side of the stream, and in addition held detached sangars high up the steep hills, and well thrown forward. Their flanks rested on very high and nearly precipitous rocks. It was only possible to climb the ridge on our right from a mile behind, and on the left from nearly three-quarters of a mile. Colonel Brander at first considered the practicability of delaying the attack on the main wall until the Gurkhas had completed their flanking movements, cleared the Tibetans out of the sangars that enfiladed our advance in the valley, and reached a position on the hills beyond the wall, whence they could fire into the enemy's rear. But the cliffs were so sheer that the ascent was deemed impracticable, and the next morning it was decided to make a frontal attack without waiting for the Gurkhas to turn the flank. No one for a moment thought it could be done.

The troops marched out of camp at ten o'clock. One company of the 32nd Pioneers, under Captain Cullen, was detailed to attack on the right, and a second company, under Captain Bethune, to follow the river-bed, where they were under cover of the high bank until within 400 yards of the wall, and then rush the centre of the position. The 1st Mounted Infantry, under Captain Ottley, were to follow this company along the valley. The guns, Maxims, and one company of the 32nd in reserve, occupied a small plateau in the centre. Half a company of the 8th Gurkhas were left behind to guard the camp. A second half-company, under Major Row, were sent along the hillside on the left to attack the enemy's extreme right sangar, but their progress over the shifting shale slopes and jagged rocks was so slow that the front attack did not wait for them.

The fire from the wall was very heavy, and the advance of Cullen's and Bethune's companies was checked. Bethune sent half a company back, and signalled to the mounted infantry to retire. Then, compelled by some fatal impulse, he changed his mind, and with half a company left the cover of the river-bed and rushed out into the open within forty yards of the main wall, exposed to a withering fire from three sides. His half-company held back, and Bethune fell shot through the head with only four men by his side--a bugler, a store-office babu, and two devoted Sikhs. What the clerk was doing there no one knows, but evidently the soldier in the man had smouldered in suppression among the office files and triumphed splendidly. It was a gallant reckless charge against uncounted odds. Poor Bethune had learnt to despise the Tibetans' fire, and his contempt was not unnatural. On the march to Gyantse the enemy might have been firing blank cartridges for all the effect they had left on our men. At Dzama Tang Bethune had made a frontal attack on a strong position, and carried it without losing a man. Against a similar rabble it might have been possible to rush the wall with his handful of Sikhs, but these new Kham levies who held the Karo la were a very different type of soldier.

The frontal attack was a terrible mistake, as was shown four hours afterwards, when the enemy were driven from their position without further loss to ourselves by a flanking movement on the right.

At twelve o'clock Major Row, after a laborious climb, reached a point on a hillside level with the sangars, which were strongly held on a narrow ledge 200 yards in front of him. Here he sent up a section of his men under cover of projecting rocks to get above the sangars and fire down into them. In the meanwhile some of the enemy scrambled on to the rocks above, and began throwing down boulders at the Gurkhas, but these either broke up or fell harmless on the shale slopes above. After waiting an hour, Major Row went back himself and found his section checked half-way by the stone-throwing and shots from above; they had tried another way, but found it impracticable.

Keeping a few men back to fire on any stone-throwers who showed themselves, Row dribbled his men across the difficult place, and in half an hour reached the rocky ledge above the sangars and looked right down on the enemy. At the first few shots from the Gurkhas they began to bolt, and, coming into the fire of the men below, who now rushed forward, nearly every man--forty in all--was killed. One or two who escaped the fire found their flight cut off by a precipice, and in an abandonment of terror hurled themselves down on the rocks below. After clearing the sangar, the Gurkhas had only to surmount the natural difficulties of the rocky and steep hill; for though the enemy fired on them from the wall, their shooting was most erratic. When at last they reached a small spur that overlooked the Tibetan main position, they found, to their disgust, that each man was protected from their fire by a high stone traverse, on the right-hand of which he lay secure, and fired through loopholes barely a foot from the ground.

The Gurkhas had accomplished a most difficult mountaineering feat under a heavy fire; they had turned the enemy out of their sangars, and after four hours' climbing they had scaled the heights everyone thought inaccessible. But their further progress was barred by a sheer cliff; they had reached a cul-de-sac. Looking up from the valley, it appeared that the spot where they stood commanded the enemy's position, but we had not reckoned on the traverses. This amazing advance in the enemy's defensive tactics had rendered their position una.s.sailable from the left, and made the Gurkhas' flanking movement a splendid failure.

It was now two o'clock, and, except for the capture of the enemy's right sangars, we had done nothing to weaken their opposition. The frontal and flanking attacks had failed. Bethune was killed, and seventeen men. Our guns had made no impression on their wall. Looking down from the spur which overlooked the Tibetan camp and the valley beyond, the Gurkhas could see a large reinforcement of at least 500 men coming up to join the enemy. The situation was critical. In four hours we had done nothing, and we knew that if we could not take the place by dusk we would have to abandon the attack or attempt to rush the camp at night.

That would have been a desperate undertaking--400 men against 3,000, a rush at close quarters with the bayonet, in which the superiority of our modern rifles would be greatly discounted.

Matters were at this crisis, when we saw the Tibetans running out of their extreme left sangars. At twelve o'clock, when the front attack had failed and the left attack was apparently making no progress, fifteen men of the 32nd who were held in reserve were sent up the hill on the right. They had reached a point above the enemy's left forward sangar, and were firing into it with great effect. Twice the Tibetans rushed out, and, coming under a heavy Maxim fire, bolted back again. The third time they fled in a ma.s.s, and the Maxims mowed down about thirty. The capture of the sangars was a signal for a general stampede. From the position they had won the Sikhs could enfilade the main wall itself. The Tibetans only waited a few shots; then they turned and fled in three huge bodies down the valley. Thus the fifteen Sikhs on the right saved the situation. The tension had been great. In no other action during the campaign, if we except Palla, did the success of our arms stand so long in doubt. Had we failed to take the wall by daylight, Colonel Brander's column would have been in a most precarious position. We could not afford to retire, and a night attack could only have been pushed home with heavy loss.

Directly the flight began, the 1st Mounted Infantry--forty-two men, under Captain Ottley--rode up to the wall. They were ten minutes making a breach. Then they poured into the valley and hara.s.sed the flying ma.s.ses, riding on their flanks and pursuing them for ten miles to within sight of the Yamdok Tso. It showed extraordinary courage on the part of this little band of Masbis and Gurkhas that they did not hesitate to hurl themselves on the flanks of this enormous body of men, like terriers on the heels of a flock of cattle, though they had had experience of their stubborn resistance the whole day long, and rode through the bodies of their fallen comrades. Not a man drew rein. The Tibetans were caught in a trap. The hills that sloped down to the valley afforded them little cover. Their fate was only a question of time and ammunition. The mounted infantry returned at night with only three casualties, having killed over 300 men.

The sortie to the Karo la was one of the most brilliant episodes of the campaign. We risked more then than on any other occasion. But the safety of the mission and many isolated posts on the line was imperilled by this large force at the cross-roads, which might have increased until it had doubled or trebled if we had not gone out to disperse it. A weak commander might have faltered and weighed the odds, but Colonel Brander saw that it was a moment to strike, and struck home. His action was criticised at the time as too adventurous. But the sortie is one of the many instances that our interests are best cared for by men who are beyond the telegraph-poles, and can act on their own initiative without reference to Government offices in Simla.

As the column advanced to the Karo la, a message was received that the mission camp at Gyantse had been attacked in the early morning of the 5th, and that Major Murray's men--150 odd rifles--had not only beaten the enemy off, but had made three sorties from different points and killed 200.

With the action at the Karo la and the attack on the mission at Gyantse began the second phase of the operations, during which we were practically besieged in our own camp, and for nine weeks compelled to act on the defensive. The courage of the Tibetans was now proved beyond a doubt. The new levies from Kham and Shigatze were composed of very different men from those we herded like sheep at Guru. They were also better armed than our previous a.s.sailants, and many of them knew how to shoot. At the same time they were better led. The primitive ideas of strategy hitherto displayed by the Tibetans gave place to more advanced tactics. The usual story got wind that the Tibetans were being led by trained Russian Buriats. But there was no truth in it. The altered conditions of the campaign, as we may call it, after it became necessary to begin active operations, were due to the force of circ.u.mstances--the arrival of stouter levies from the east, the great numerical superiority of the enemy, and their strongly fortified positions.

The operations at Gyantse are fully dealt with in another chapter, and I will conclude this account of the opposition to our advance with a description of the attack on the Kangma post, the only attempt on the part of the enemy to cut off our line of communications. Its complete failure seems to have deterred the Tibetans from subsequent ventures of the kind.

From Ralung, ten miles this side of the Karo la, two roads branch off to India. The road leading to Kangma is the shortest route; the other road makes a detour of thirty miles to include Gyantse. Ralung lies at the apex of the triangle, as shown in this rough diagram. Gyantse and Kangma form the two base angles.

[Ill.u.s.tration]

If it had been possible, a strong post would have been left at the Karo la after the action of May 6. But our small force was barely sufficient to garrison Gyantse, and we had to leave the alternative approach to Kangma unguarded. An attack was expected there; the post was strongly fortified, and garrisoned by two companies of the 23rd Pioneers, under Captain Pearson.

The attack, which was made on June 7, was unexpectedly dramatic. We have learnt that the Tibetan has courage, but in other respects he is still an unknown quant.i.ty. In motive and action he is as mysterious and unaccountable as his paradoxical a.s.sociations would lead us to imagine.

In dealing with the Tibetans one must expect the unexpected. They will try to achieve the impossible, and shut their eyes to the obvious. They have a genius for doing the wrong thing at the wrong time. Their elan, their dogged courage, their undoubted heroism, their occasional acuteness, their more general imbecile folly and vacillation and inability to grasp a situation, make it impossible to say what they will do in any given circ.u.mstances. A few dozen men will hurl themselves against hopeless odds, and die to a man fighting desperately; a handful of impressed peasants will devote themselves to death in the defence of a village, like the old Roman patriots. At other times they will forsake a strongly sangared position at the first shot, and thousands will prowl round a camp at night, shouting grotesquely, but too timid to make a determined attack on a vastly outnumbered enemy.

The uncertainty of the enemy may be accounted for to some extent by the fact that we are not often opposed by the same levies, which would imply that theirs is greatly the courage of ignorance. Yet in the face of the fighting at Palla, Naini, and Gyantse Jong, this is evidently no fair estimate of the Tibetan spirit. The men who stood in the breach at Gyantse in that h.e.l.l of shrapnel and Maxim and rifle fire, and dropped down stones on our Gurkhas as they climbed the wall, met death knowingly, and were unterrified by the resources of modern science in war, the magic, the demons, the unseen, unimagined messengers of death.

But the men who attacked the Kangma post, what parallel in history have we for these? They came by night many miles over steep mountain cliffs and rocky ravines, perhaps silently, with determined purpose, weighing the odds; or, as I like to think, boastfully, with song and jest, saying, 'We will steal in upon these English at dawn before they wake, and slay them in their beds. Then we will hold the fort, and kill all who come near.'

They came in the gray before dawn, and hid in a gully beside our camp.

At five the reveille sounded and the sentry left the bastions. Then they sprang up and rushed, sword in hand, their rifles slung behind their backs, to the wall. The whole attack was directed on the south-east front, an unscalable wall of solid masonry, with bastions at each corner four feet thick and ten feet high. They directed their attack on the bastions, the only point on that side they could scramble over. They knew nothing of the fort and its tracing. Perhaps they had expected to find us encamped in tents on the open ground. But from the shallow nullah where they lay concealed, not 200 yards distant, and watched our sentry, they could survey the uncompromising front which they had set themselves to attack with the naked sword. They had no artillery or guncotton or materials for a siege, but they hoped to scale the wall and annihilate the garrison that held it. They had come from Lhasa to take Kangma, and they were not going to turn back. They came on undismayed, like men flushed with victory. The sepoys said they must be drunk or drugged. They rushed to the bottom of the wall, tore out stones, and flung them up at our sepoys; they leapt up to seize the muzzles of our rifles, and scrambled to gain a foothold and lift themselves on to the parapet; they fell bullet-pierced, and some turned savagely on the wall again. It was only a question of time, of minutes, and the cool mechanical fire of the 23rd Pioneers would have dropped every man. One hundred and six bodies were left under the wall, and sixty more were killed in the pursuit. Never was there such a hopeless, helpless struggle, such desperate and ineffectual gallantry.

Almost before it was light the yak corps with their small escort of thirty rifles of the 2nd Gurkhas were starting on the road to Kalatso.

They had pa.s.sed the hiding-place of the Tibetans without noticing the 500 men in rusty-coloured cloaks breathing quietly among the brown stones. Then the Tibetans made their charge, just as the transport had pa.s.sed, and a party of them made for the yaks. Two Tibetan drivers in our service stood directly in their path. 'Who are you?' cried one of the enemy. 'Only yak-drivers,' was the frightened answer. 'Then, take that,' the Tibetan said, slashing at his arm with no intent to kill. The Gurkha escort took up a position behind a sangar and opened fire--all save one man, who stood by his yak and refused to come under cover, despite the shouts and warnings of his comrades. He killed several, but fell himself, hacked to pieces with swords. The Tibetans were driven off, and joined the rout from the fort. The whole affair lasted less than ten minutes.

Our casualties were: the isolated Gurkha killed, two men in the fort wounded by stones, and three of the 2nd Gurkhas severely wounded--two by sword-cuts, one by a bullet in the neck.