One of the least known mountain ranges of the Indian border is that which forms the watershed between the Lohit-Brahma- putra and the upper Irrawaddy, where Assam, Burma, and Tibet meet.1 It is true that only the southern end of this range is in British- Indian territory; but since farther north it blocks what is virtually the only way into India between Burma and Tibet, it is of considerable significance. So far as appearances go, this Lohit-Irrawaddy divide is only a southern arm of the mightier Tsangpo-Brahmaputra- Salween divide—the so-called Ninchin-Thang-La range; the other branch forming the divide between the Irrawaddy and the Salween. But appearances may be deceptive. All the rivers here run from north to south; the mountain ranges between them therefore appear to do the same. But, as is well known, farther north, all the rivers flow from west to east, or at any rate south-eastwards, and again the mountain ranges between them appear to—and in fact do —trend east and west. Therefore, unless the mountains definitely curve round southwards, they must belong to two quite distinct systems.
There is, of course, a third possibility, namely, that the north- south ranges represent the flanks of grooves cut in the main east-west ranges by rivers flowing through them; but this is discounted by the fact that geologically the north-south ranges appear to be features of original structure and older than the Himalayan ranges. However, this very circumstance makes it certain that they must have been profoundly affected by the uplift of the much greater Himalayan ranges. Indeed, the rather abrupt elevation of these meridional ranges along the Burma-China frontier, north of latitude, say 270, suggests that these snow peaks owe their presence, not to the original formation of these ranges, but to a secondary impetus due to the uplift of the Himalaya. In other words, if we could plot all the snow peaks north of the 27th parallel, between the meridians of 96° and 1oo°, we should find that they really trend east and west, across the rivers and across the grain of the country, and not on short meridional lines, as they appear to the traveller to do. Although, owing to the peculiar river system, they appear to belong to the older uplift, they actually belong to the newer.
Footnote
1 The sketch-map (p. 48) is published by permission of the Royal Geographical Society. See also Survey of India maps 91 H and 92 E.
To return, then, to the Lohit-Irrawaddy divide. If the above theory is correct, it is not a question of the Ninchin-Thang-La range (an east-west range) dividing into two at its eastern end, with the Irrawaddy rising in the angle between them; but of that river rising from the southern slope of the range and gnawing its way back into it, leaving great spurs on either flank. To the west of one spur the Lohit flows southwards; to the east of the other the Salween does the same.
Confining our attention to the Lohit-Irrawaddy divide, as we may continue to call it,9 this range stretches southwards, or south- westwards, from the source of the Irrawaddy in latitude 28°3o' N. to about 27°3o' N. before bifurcating again into the Patkoi and Kumaon ranges. In the course of its seventy odd miles it is crossed by five passes, named, from south to north, Hpungan pass (10,080 feet), Kumjawng pass (9,682 feet), the Diphuk La (14,280 feet), the Shori La (height unknown), and the Zasha La (15,711 feet).10
At its southern end the range is low, the highest peaks in the neighbourhood of the Hpungan pass barely exceeding 13,000 feet, while in the immediate vicinity of the Kumjawng pass they hardly reach 12,000 feet. A little farther north they jump up to 15,000 feet again, while in the neighbourhood of the Diphuk La they exceed 16,000 and even 17,000 feet. Between the Diphuk La and the Shori La there is again a sudden increase in elevation above the snow-line to over 19,000 feet, while around the Zasha La also there are lofty peaks.
Footnote
It was to explore again on the Lohit-Irrawaddy divide that Lord Cranbrook and I visited Upper Burma in 1930-1. I had crossed the Diphuk La in 1926; but we proposed to explore farther north. Our objects were to collect fauna and flora at the sources of the Irrawaddy and to continue the tracing of the southern limits of the great ice- sheet which once covered so much of this country from the Tsangpo to the Yangtze and beyond.
We arrived in Burma in November 1930, and left Myitkyina, the northern terminus of the Burma railway, on the 26th. Using mule- transport we reached Fort Hertz (214 miles) on the 16th December. There is now a dry-weather cart-road from Myitkyina to Sumpra Bum, rather more than half-way to Fort Hertz. Motor-cars even have done the journey, but it would be flattery to call it a motor- road, since a day's rain is sufficient to render much of it impassable. The first fifty miles is, however, regularly covered by lorries and buses, of which there are two or three marvellous antiques in Myitkyina, owned, of course, by Chinese contractors. It is only aquestion of time—and money—before there will be a very fair dry- weather dirt-road to Sumpra Bum, and there is no insuperable obstacle to carrying such a road on to Fort Hertz. To be able to reach Fort Hertz in three days from Myitkyina would be a great step forward; but lorry hire would be expensive, though not more so than that of mules. At present the charge for a mule is Rs. 30, as payment must be made for the return journey as well as the outward one; and a mule only carries 120 lb., the trip taking nineteen days.
Sources of Irrawaddy
Looking north up the Nam Tamai towards the Tibetan frontier range
After halting for several days at Fort Hertz we set out for the Nam Tamai, which lies to the north-east; the Tamai is the second largest branch of the eastern Irrawaddy. From Fort Hertz to the Tamai is nine short, but steep, stages, soon to be reduced to eight. Owing to the attitude of the Chinese authorities, or rather of the Yunnan authorities, who have laid an embargo on the activities of the Chinese mules working east of the Myitkyina-Fort-Hertz road, we could not take mules to the Tamai, though there is a dry-weather mule-path (not a cart-road) all the way. So we had to use coolie transport. The difficulty could have been overcome by engaging mules from the Shan States instead of Yunnan animals in Myitkyina; but it was hardly worth the expense, since we should have had to pay for the mules, not only returning empty all the way, but also for the journey to and from the Shan States as well. For our part we were in no hurry after leaving Fort Hertz; but it is well to record that we could have procured mules in Myitkyina which would have gone right through to the Tamai river with us, and possibly several marches up that stream.
Travellers for Fort Hertz must remember that they can buy hardly anything on the road, though there are one or two Chinese shops at stated intervals, where tinned milk (mostly of an ancient vintage!), kerosine, and other luxuries can be obtained, at a price. There are half a dozen shops in Fort Hertz itself, where tinned stuff in some variety can be bought; and what perhaps is more to the point, unlimited rice. Between Fort Hertz and the Tamai nothing at all can be bought, though the road passes through one village, Nogmung, four marches from the edge of the plain. Neither can any supplies be obtained in the Tamai valley, up which we marched for six days' journey—-actually not much more than fifty miles—in January 1931.
By the end of the month we had reached the Seinghku-Adung confluence. I had explored the Seinghku valley in 1926, and crossed the Diphuk La at its head to the Lohit river twice. The Adung river is the main stream, that is, the upper course of the Tamai; and it was this branch that we followed in February 1931, using the local Daru coolies for transport. These are the people variously called Taron, Tellu, and other names, or at any rate other pronunciations. There is some reason to believe that they are connected with the Mishmi of Assam; anyhow there are very few of them, and they are shockingly oppressed by the Tibetans, Lisus, Chinese, and indeed by every one who comes into this country, except of course the British. The natural result is that the British are the one people who cannot make them do anything. They will work, willingly or unwillingly, for Tibetans, or Lisus, or any one else; but they will not work for the British.
The problem of travel in the upper Irrawaddy jungles is entirely a problem of transport. Everything has to be carried, because nothing can be obtained locally. The valleys at the sources of the Irrawaddy are not entirely uninhabited; but one might easily think that they were, for all one sees of man or hut. Such villages as there are do not remain in the same place for more than two or three years at the most; the tribes, if not nomadic, are at any rate vagrant.
We reached the Tamai on the 7th January, the Seinghku confluence (six stages) on the 25th, and Tahawndam (four stages) on the 5th February, though our last loads did not arrive till the 12th. In all we had about seventy loads, so that the rate of progress was not startling. In the rainy season, if possible at all, it would have been very much slower; but it is hardly possible to move about, below 10,000 feet, during the rains, and above that altitude it is impossible to travel between December and May because all the passes are then blocked by snow. Movement at the headwaters of the Irrawaddy is therefore restricted.
Tahawndam is literally the last village in Burma; but it is more Tibetan than Daru. In the Tamai valley there are mithan, but no other cattle; in the Adung valley there are not only cattle and half- breed yak, but also sheep and goats, as in the Seinghku valley. This is a sure indication of a more temperate climate; and indeed Tahawndam is not only at 6,000 feet altitude, but it is completely surrounded and overshadowed by hill ranges whose peaks attain 15,000 feet or more and are snow-clad for at least nine months of the year. Our ultimate object was to cross the pass at the head of the Adung valley into Tibet; that is, if there was a pass. We were told that there was, and moreover that it was regularly used. In fact, as we found, the people of the Adung valley traffic with the Tibetan villages over the Namni La, and not with the people of the Tamai and Fort Hertz.
At Tahawndam we found permanent cultivation and grazing (on south slopes), as well as the usual shifting hill-side cultivation of the Darus. The entire population of the valley and its tributaries above the Seinghku confluence may number 500 men, women, and children. No doubt it varies considerably from year to year according to the incidence of famine, pestilence, and raiding; it has probably never exceeded that figure. Consequently the difficulty of assembling even thirty or forty coolies at one place on the same day may be imagined. Actually it cannot be done; even twenty are rarely obtainable, and usually for ten one must be thankful. Probably half of these will have insufficient food for a two-days' march, though they require little enough in all conscience. Any one who contemplates exploring or collecting in this country should arrange to dump a hundred bags of rice—double-bagged and padlocked— at a forward village during the so-called dry weather, and be prepared to take on and feed permanently ten coolies, preferably bringing them with him from another part of the country.
Daru rope-bridge over the Nam Tamai just below the Seinghku river confluence
As it was we made our next move up the valley, three days' march, to 8,000 feet in the middle of May. In June we ascended to our alpine camp at 12,000 feet, and there, three hours' climb from the pass, we spent the summer. In September we crossed the Namni La which leads, not directly over the Lohit-Irrawaddy divide, but over a lofty range separating two of the Irrawaddy headwater streams, the Tamai and the Taron.
On either side of the pass we observed five glaciers; there is a sixth a little farther south, and there are almost certainly one or two more to the north-east though we could not see them. They are all either hanging or corry glaciers; and besides these there are also several snow-beds. The group of glaciers round the Namni La, however, are not the most southerly in Burma; so far as my own observations go, there are no glaciers in the Seinghku valley or south of the Diphuk La, but there may be a few small ones north of that pass, before the 19,000-foot peaks are reached. It is probably correct, however, to say that there are no glaciers south of the 28th parallel, between the Brahmaputra and the Salween to-day; and only remnants south of latitude 290, or even 30°, in that area. Except for the Burma-Assam re-entrant, between meridians 940 and 96°, the southern edge of the ice-sheet certainly came down to the 28th parallel, and still farther south in the Himalaya and in China. It is important to notice that this ice-sheet at one time completely cut off the Indo-Malayan region from the central Asian region, and the central Asian region from the eastern Asiatic; but it never completely cut off the Indo-Malayan region from the eastern Asiatic, which, farther east, must have long been in contact for hundreds of miles.
Tahawndam, the last village of the Adung valley, is a Tibetan settlement comprising three families. The Tibetan system of cultivation is very different from that of the jungle tribes. It is always permament; the Tibetans cannot live except where permanent cultivation is possible, so it is unlikely that they will push any farther down the Irrawaddy. As it is, 6,000 feet is very low for them. The principal crop at this altitude is barley. The Tibetan also requires permanent grazing; and we found the southern slope of the valley denuded of forest, covered with high grass and bracken. It is burnt afresh every year about March.
There were herds of yak or half-breed yak, goats, and sheep at Tahawndam, and up to the time we left in the middle of May, the Tibetans supplied us with fresh milk every day, and with butter weekly; we had enough for our needs, but of course could have done with more, had it been available. After we went up the valley, we got no more milk and only a little butter occasionally. There was not enough grain cultivated to supply us; but from time to time we obtained a little tsamba (roast barley-flour), Indian corn, peas, wild honey, and other things. For meat the Tibetans relied largely on pork from the village pigs, chickens, and on the chase—serow, gooral, and barking-deer are all common. The Tibetan hunts with dogs and short gas-pipe guns; the Daru only uses the cross-bow, with poisoned arrows, but he also snares animals and pheasants. About June the Tibetans take their animals up the hill-side to the higher pastures at 8,000 feet; the forest on southern slopes having first been burnt in some places up to nearly 10,000 feet. There is a good path from the valley on to a shoulder, whence the path ascends the ridge into the Abies forest. I followed it to over 11,000 feet, and there is no doubt that in the summer the cattle go right up to the alpine region, immediately overlooking the village. By this time, however, we were far up the valley.
The mountains are composed entirely of granite and are precipitous. Enormously high smooth cliffs are exposed, and the streams fall into the river in a series of leaps. Washouts and rock-avalanches are of frequent occurrence. Three small streams, dry in winter, which descend close to the village from a common source at about 14,000 feet, came down in spate during the heavy rains of July and did a good deal of damage. Thousands of tons of gravel and rock were washed down, obliterating a field and the site of our camp, but just missing the village. There is not much doubt that these washouts are aggravated by burning the forest on the south slopes. Nevertheless, owing to the greater accumulation of snow on the north face destructive avalanches and washouts occur here also, and there is little to choose between them.
I have mentioned wild honey above. In the months of May and June this honey is obtained largely from the many rhododendrons which grow in these forests and is poisonous, at least to some people. Classical students will remember the sad story of the Retreat of the Ten Thousand in Xenophon's Katabasis and how the Pontine honey affected the Greek soldiery. Cranbrook, having eaten his first meal of wild honey at Tahawndam, went off into a trance, and two of our native servants were ill. I was not affected, possibly because I ate more moderately of the 'Tree of Knowledge'. But it is a point well worth bearing in mind.
Honey was not the only eatable we found wild in the Adung valley. Bamboo shoots and a bright orange-coloured fungus which grew in enormous tiers on the silver firs at from 11,000 to 12,000 feet afforded a welcome supply of fresh food (vitamin content unknown!) in the summer; also there were in the alpine meadows unlimited supplies of a very mild garlic (Allium sp.) of which we ate the leaves chopped up. These things proved very useful during the three months when we had no potatoes. Big game—if gooral can be classed as big game—having proved itself unprocurable at about 8,000 or 9,000 feet up the valley (Cranbrook shot two gooral on the opposite cliff from our camp, but we failed to recover the bodies), and nonexistent in the alpine camp at 12,000 feet, we relied on smaller game for fresh meat; squirrels, pheasants, and even smaller birds were all welcome to the pot.
In July visitors from Tibet came over the top into our valley and asked—demanded would be a better word—permission to dig up pai-mu; this is the Chinese word for the bulb of Fritillaria Roy lei, a well- known general tonic in China. The 'rights' in this valley and its main tributaries apparently belong to a wealthy Tibetan family which lives in the Salween valley two days' journey south of Men- kung; I stayed in their mansion in 1913 and can testify to their wealth. The Tibetan headman—I believe he is even a petty official —himself came over the pass; but for the actual work he employs a gang of ruffians from the Tzekow region on the Mekong—Sino- Tibetans, Lisus, and other banditti. The Tibetan spoke nothing but Tibetan; but the Chinese interpeter spoke Daru, as well as Chinese and Tibetan; and it was with him that I negotiated. I had the not very bright idea of formally giving these unpleasant-looking bandits permission to dig pai-mu in the Adung valley, in return for their taking us over the pass and entertaining us hospitably on their side, and even allowing us to travel northwards into drier country. This scheme was only partially successful. When the truculent Tibetan chieftain had gone back to his own country, the Mekong contingent of semi-Chinese warriors did indeed return to us and carried ten loads for us three days' journey to the first village the other side of the pass, called Jite. But beyond that they would not budge, saying that they were afraid; and the Jite people said likewise. This was why the Tahawndam folk would not take us over the pass either. It was a pity we could not reach Ridong on the other big branch of the Taron, said to be three marches from Jite, over three mountain ranges; these 'ranges' could only have been spurs, separating southward-flowing tributaries of the Jite river.1 Both Cranbrook and I eventually reached the top of the first spur above Jite, but owing to the bad weather we could see nothing. However, the main thing was to cross the pass, and this we accomplished.
Footnote
1 See Map, Geographical Journal, vol. lxxx, p. 474, from Lord Cranbrook's compass traverse.
With half a dozen coolies in our permanent employ we could have easily reached Ridong, but the weather throughout September was so bad that, from a mapping point of view, it would have been little use. So far as we can judge, Ridong seems to be placed on existing maps several miles too far north; Bailey did not get an observation for latitude there and no other European seems to have visited it. I was told there was a good cantilever bridge over the Taron at Ridong, and about thirty houses scattered about. There is said to be no path down to the river, near which there are only a few scattered wild Tarons (Darus). Nevertheless Daru slaves are fairly numerous in the Tibetan villages upstream and as far as the Salween river. Perhaps some day Geneva will interest itself in the matter—or is it too remote? Meanwhile I may remark that Lhasa is unable to make its voice heard in this part of Tibet; for I was assured, locally it is true, that the Tibetan Government had strictly forbidden its nationals to enter the Adung valley, pending a decision as to whether the Indian Government proposed to claim or abandon it. All this makes it rather doubtful whether a Tibetan passport would have been of any assistance to us.
It was, however, a satisfaction to us to cross the Namni La on the 1 st September and to continue our observations on the glaciation of this region. The river we followed down to Jite has its sources in several glaciers and is one of three head-waters which together make up the Taron river. Possibly it is the largest, and, if so, then we found what is technically the source of the Irrawaddy—though we claim no special merit for that! We are satisfied to have found one of its many sources.
The climate on the other side of the range is distinctly colder and drier than on the Burmese side, but we derived small satisfaction from this fact in September. It rained practically the whole of the three weeks we were there, and when not actually raining, we could see nothing for mist. That the climate was not quite the same was, however, proved by a different type of vegetation, as well as by the appearance of different species. There was also the fact that the last human habitations in the Adung valley are at 6,500 feet, while Jite is about 4,000 feet higher, a fact which may, however, be due partly to the nature of the valley. At any rate there is no gainsaying the point that the Adung valley is in direct communication with the dense sub-tropical forest region of northern Burma; whereas the Jite valley is linked with the chilly wind-swept grass plateau region of eastern Tibet.
On the other hand, if the grassland plateau region is drier than the forest region to-day, one suspects that it was not always so. The north side appears to be far more glaciated than the south side; and, as already stated, there are several glaciers on the north side, and only two insignificant ones on the south. It is true that the drier climate might preserve the effects of ice-action better; but it could hardly initiate larger glaciers. On the southern slope, the wetter climate might destroy the evidences more effectively, but it would keep the glaciers in action longer. Some argument may therefore be made out for a gradual change of climate in this part of Asia.
Looking up the Adung river from Base Camp at 6,000 feet, in February
It is worth noting how very sparsely inhabited this mountain country between the Lohit and the Irrawaddy really is. Grossing the range by the Diphuk La, it is eight good marches from village to village, or eleven marches from the Nam Tamai to the Lohit, passing perhaps four villages, comprising in all from twenty to twenty-four houses, with a total population of a hundred souls. Grossing by the Namni La, it is seven long marches from village to village. Reduced to miles, the distances may not sound formidable, but any one who thinks that a seven-mile march up one of these valleys is not an honest day's toil, had better try it!
After crossing the Namni La on the 1st September we descended to a lovely little alpine lake, of which there are many scattered about in these mountains. Below the lake the valley ended in a cliff, perhaps 500 feet high and overlooking a larger valley, the upper part of which was, as usual, flat; farther down, this valley ended also, not exactly in a cliff, but in a very steep ravine, which led to a yet larger valley. In this, the main valley, there was an enormous pasture, probably 300 hundred acres in extent, where yaks, sheep, and goats grazed. There was no permanent habitation here, at an altitude of 12,500 feet, but from here onwards the path was fit for pack animals. Farther down we crossed a fairly big river from the north; this was the Shori river, and at its head there is a pass over the Lohit-Irrawaddy divide, carrying the road to Sanga-chu-tzong and Lhasa. There may be villages up this valley—certainly there are a few families. A cantilever bridge spans the Shori river, and on the other bank is a shelter hut, evidently a favourite halting-place for the night. From here to Jite is a short day's march.
Arrived at Jite on the 4th September, we found the local inhabitants reaping the barley. It is a miserable village of three or four houses, standing on a bare shoulder of the mountain, over 10,000 feet above sea-level. A considerable river flows in from the north, in a deep valley, to join the main river just below the village. The combined river turns more and more towards the south, and joins the Ridong river (Kalaw Wang) at some unknown point.
The Jite river flows in a very narrow gorge, shut in by sheer cliffs, and I was unable to reach it, much less to follow it. The road to Ridong, a very good one, though muddy, diverges from the river, keeping more to the south-west. I followed a good path up the northern branch of the Jite river for several miles, and came to another house, but the path ended there, and I was told that there was no route up the valley, a piece of information I found difficult to credit, though I could not disprove it.
In this part of Tibet the valleys are well forested up to 12,000 feet —juniper and fir above, birch, maple, oak, Pyrus, Prunus, and many other trees below. There is abundant meadow and grazing country. The climate is severe. The people speak a bastard kind of Tibetan, but have few dealings with Tibet proper, and perhaps fewer with Yunnan. No doubt the sturdy folk of the Mekong valley 'milk' them as they in turn 'milk' the people of the Irrawaddy jungle.
This part of Tibet appears to have no comprehensive name. It is not part of Zayul, nor is it part of Tsarong. Zayul is presumably a province; Tsarong a smaller designation, a district, comprising the arid part of the Salween, around Menkong. The name Tsarong means 'the hot valley'. The Jite district lies to the south-east of Zayul, and south of Tsarong; it is certainly not hot, and still less dry.
The Jite people were very frightened of us. Translated, that simply means that they were frightened of incurring the displeasure of their officials, owing to our presence. But greed triumphed over policy. When they found that we had money, and when their officials demanded taxes, they came to us with offers of food and raiment. We bought a little tsamba and cornflour, also butter, potatoes, and turnips. Had we reached Ridong we might have bought unlimited supplies; as it was we did not get very much, but enough to go on with.
The officials and 'the church' would have no dealings with us. I was told there was no official at Ridong; but there is no doubt that news of us had spread all over Tsarong, and perhaps Zayul also. There was a priest of sorts—not a lama in the strict sense—at Jite; but he kept well out of our way and had not even the courtesy to call. The general boycott made me wonder whether we should ever get coolies to take us back to Burma; we might just be abandoned and forgotten, I thought. I therefore sent our only Daru coolie back to our own village with a message to the headman, requesting him to send a rescue party. If ever he received the message, he did nothing about it.
However, after twenty days, the Tibetans collected some ponies and took us back to the foot of the pass. Though we had only ten or twelve loads, we had only five or six coolies, and as soon as the transport animals had to be abandoned we had to move in relays. We recrossed the Namni La under depressing conditions, pouring rain, a bitterly cold wind, and thick mist. Arrived at our camp, the Tibetans departed and we were stranded again, out of touch with either village. Cranbrook sportingly volunteered to march down to Tahawndam for help, and he accomplished the journey in five days, while I stayed in the 12,000-foot camp and collected seeds. At the end of October, after three weeks in the alpine camp, I rejoined Cranbrook at the next camp below, to which he had come back; and we leisurely returned to the village, collecting as we went. By the first week in November we were all back at the base camp making preparations for departure. We left Tahawndam on the 20th November and were back in Fort Hertz early in December. Christmas was spent on the road, and we reached Myitkyina on New Year's Day 1932, after an absence of just over thirteen months.
The natural history results of the expedition were very good. Cranbrook made an excellent collection of birds and mammals, including several new species; the range of other species was extended and more clearly defined. We also collected a number of insects, including new and rare species. Though our half-dozen snakes included no new ones, the small collection proved quite interesting to those who like such things. The botanical collection requires more working out, but it includes many new species and some beautiful plants which have been introduced into cultivation.
The geographical results are less satisfactory, for we were held up sooner than I had anticipated. In order to travel even in these remote parts of Tibet without permission from Lhasa—or even perhaps with it—two conditions are absolutely essential. Firstly, one must have a nucleus of reliable coolies, drawn from some other part of the country, on whom the Tibetans cannot visit reprisals, and who must be, therefore, dependent on one for food; and secondly, one must move with speed. Had we been able to feed ten coolies, and to cross the Namni La with our own men before the news of our arrival at Tahawndam reached Jite, there would have been nothing to prevent us from reaching Ridong at any rate, and perhaps Tsarong. But obviously even this method presently breaks down. The official Tibetan reply to these tactics is to place an embargo on all food; any one who sells food to the unpopular party does so at his peril. Ten coolies cannot carry loads and enough food for themselves for very long.
Nevertheless, our glacial observations were decidedly interesting and satisfactory; and we discovered this route directly connecting Burma with Tibet. Now, having crossed both passes, I unhesitatingly pronounce in favour of the Diphuk La, as a direct link.