Himalayan Journal vol.07
The Himalayan Journal

Publication year:

Kenneth Mason
    (H.W. Tilman)
    (I. GENERAL r. finsterwalder)
  5. A VISIT TO NUN KUN, 1934
    (F. S. SMYTHE)
    (J. B. AUDEN)
    (A.P.F. Hamilton)
    (Captain F. KINGDON WARD)
    (P. C. DUNCAN)
    (Lieut.- Colonel Kenneth Mason)
  16. NOTES



(Translated from the German by Lieut.-Col. C. M. Thompson, i.a.)[1]
The Second German Kangchenjunga Expedition (1931) included among its projects the survey and preparation of a map of the country traversed. The earlier expeditions under Freshfield (1899) and Dyhrenfurth (1930) had made maps of these parts, and the Survey of India sheets had been prepared as fine general maps showing the broader features of the wild, branching, mountain chains in the neighbourhood of Kangchenjunga. Nevertheless no really reliable map existed, entirely trustworthy as regards the minor details. To prepare such a map was the object of our enterprise. The difficulties of such a task are considerable, for the Himalaya, with their rugged shapes, icy and avalanche-swept slopes, peaks, glaciers, and ravines, were bound to present formidable obstacles. There was no doubt about the best method of survey, for it is only by modern stereo-photogrammetry that it is possible to survey the higher features of the mountains exhaustively and give a true representation of their form.

Of the two types of photogrammetry, from the ground and from the air, the latter was impracticable, not only because of expense, but also owing to the difficulty of flying high enough above the objective, itself over 8,500 metres. Thus terrestrial photogrammetry was the best method that remained. I had already found that this was a sound method in the Alps and on the Pamirs, where I had learned something of its difficulties in wild and inaccessible regions. I was therefore at first inclined to advise Bauer not to hamper his expedition with such a task when his main object was to climb Kangchenjunga itself, an object on which his whole energies should be concentrated.

The field-work of such a photo-survey cannot be carried out from paths in the valleys, nor by climbing to the top of a single peak. It involves climbing to a number of suitable view-points throughout the country and resolves itself more or less into a mountaineering conquest of the whole region. In this work, technical efficiency, bodily fitness and energy, and mountaineering knowledge, all play important parts. For these very reasons I felt that it was incumbent on a mountaineering expedition such as Bauer's to take up the project, and I resolved to undertake the photogrammetry of the Zemu glacier and the neighbouring mountains.

It was lucky that we had in Karl Wien an expert in the field-work. He had already carried out a similar project on the Pamirs. It is thanks to him that the Zemu glacier map appears as the result, for the field-work was brought to a conclusion by him with energy and resolution even after the formidable struggle with Kangchenjunga had ended.

The map that Bauer planned was first of all to be an accurate and comprehensive chart of the routes followed by the expedition. It was also necessary to ensure that the heights should be thoroughly reliable.1 Ordinary photographs and verbal descriptions can never illustrate the structure and elevation of such a country so satisfactorily as a map, which for this reason becomes an important part of any mountaineering project. An accurate map is also of high scientific value for geography, glaciology, and geology, especially in this instance, for it constitutes the first accurate large-scale survey of a great Himalayan glacier and the first detailed reproduction of a limited portion of the high Himalaya, ‘young' from the point of view of geology, and rich in the varied formations it displays.

Existing maps and data.

Photographic survey is so much better and more complete for such a country than older methods, that previous maps, not based on the method, were not of much value for the preparation of our new map. Nevertheless, for many reasons, the previous maps were of interest to us. They are the following:

1. The Survey of India map, scale 1 : 253,440. This covers the whole of India, and our area falls in sheet 78A, between latitudes 270 and 28°, and longitudes 88° and 89° east of Greenwich. It is mainly reliable where it covers inhabited areas,2 and it gives the general lay-out of the mountains correctly; but the representation of the country in this part is sketchy, with contours at 500-foot intervals, glaciers are merely indicated, and only the main peaks are correctly shown with positions and heights

1 Height measurements at great heights and at points far from meteorological stations, formerly mostly derived from barometric observations, can easily be incorrect by hundreds of metres.

2 But there are also errors, e.g. in the upper Talung valley places are shown, whereas actually west of Sakhyong the country is an uninhabited wilderness, as pointed out in 1931 by Allwein and in 1920 by Tobin.
  1. The 'Sketch-map of the Glaciers of Kangchenj unga, by E.J. Garwood, scale 1 : 125,000. This map was prepared on Freshfield's expedition of 1899, and for our region shows a considerable improvement on the Survey of India map. Greater importance was attached to the representation of mountain features, as Freshfield pushed forward on all sides towards Kangchenj unga. Garwood visited a large number of high points, which are distributed throughout the area, and made a rapid survey from them. He also took several photographs. From these the map was prepared at home.[2] The map shows the lie of the main mountain ridges and also their branches towards the smaller side ranges in a generally accurate and complete manner. The glaciers, including the side ones, are depicted with particular care and are accurately mapped in position and size; their features, such as moraines and crevasses, are shown; and although the map is somewhat lacking as regards being an exact representation of the ground-e.g. it contains only rough shading and no contours-yet it will always be a good example of what a general map should be, and in many respects even now it can serve as a model for methods of survey and representation of features. It has proved itself complete and reliable in the Zemu area within the limits prescribed for it.[3]
  2. The massif of Kangchenj unga, scale 1 : 100,000, by Marcel Kurz.[4] This map was the result of the International Himalaya Expedition in 1930, and has improved on the Garwood map in the area west and north of Kangchenj unga, where the expedition first worked. Heights are shown by contours. The well-trained hand of the Swiss cartographer Kurz is traceable in the artistic, if somewhat scanty, representation of the rock features. In the Zemu area, where this expedition took no photographs, but used mainly those of the Bauer expedition of 1929, the map shows practically nothing new.
Far more important than the foundation provided by these maps were the triangulation data of the Survey of India, which has fixed the most important peaks in our area from the triangulation chain at the foot of the Himalaya, and published the co-ordinates. Four important starting-points were taken from Triangulation Pamphlet for Sheet 78A, published by the Geodetic Branch, at Dehra Dun in 1918.

Outline of photogrammetric method.

Photogrammetry serves the double purpose of triangulation and detailed topographical survey.

Triangulation, which forms the basis of every accurate map, especially those to be made by stereo-photogrammetric methods, is usually carried out with a theodolite. Suitable triangulation stations are selected over the whole area, from which reciprocal observations are taken, so that the sides uniting them form a net of approximately equilateral triangles. The stations are carefully marked with signals, and the angles observed, so that the positions of the stations can be calculated. It is a lengthy and expensive business, which has to be carried out after a careful reconnaissance, and it is difficult for an Alpine expedition to use it.

A way out of the difficulty can be found by photogrammetry, as was done in the case of the Zemu survey. For if panoramas are taken of the triangulation stations, selected in other respects as before, the angles of the net can be derived from the photographs at home, and the measurement of the angles in the field, a laborious part of the work, can be avoided. It is only necessary to set up the camera on well-marked points, so that the positions can be identified on the photographs. If the points are judiciously selected, signals need not be erected.

More important than this supplementary triangulation is the actual survey, for which the method was specially introduced. In principle this is quite simple. At favourable points, with a good view of the country, stereo-photos are taken, which, if possible, cover the whole area, with no gaps.

These photographs are different from ordinary stereoscopic photographs, owing to the distance between the points of view-the base- being very much greater than the distance between the eyes. The length of the base should be from 1 /1oth to 1 /20th the distance from the object photographed, e.g. to photograph a slope 5 km. distant, the base should be between 250 and 500 m. The directions in which pairs of photographs are taken are also generally parallel, but not always perpendicular to the base, and sometimes inclined as much as 30° to it. The length of the base should be measured accurately, generally by means of an auxiliary base and simple triangulation.

The camera axes should be parallel, or convergent at a known angle. At least one of the points from which a pair of photographs is taken should have its position determined, which can be done by means of a panorama taken from it and including recognizable fixed points in the panorama. In order to attain successful results, a special form of camera, known as the photogrammeter, provided with compass, is necessary.

The photographs are taken in succession by means of this instrument at both ends of the base. About two hours of strenuous work are necessary for the observations. The base stations must first be clearly marked by signals, so that the reciprocal rays can be observed within a minute of arc. This is done by the erection of small cairns over which the instrument can be set up in turn. The measurement of an auxiliary base, about 1 / 10th of the length of the base-line, follows, after which the taking of the photographs and some compass readings are necessary in order to determine the position.

It is not always possible to set out the base-lines on easily accessible positions, as the most favourable panoramic positions are mostly on high ridges, necessitating laborious and perhaps perilous ascents. One may have to work on dizzy eminences in spite of cold and fatigue, and yet carry out the work and record the observations carefully. Any inaccuracy or error may ruin the results. The observations must be taken at the most favourable time of day, when light is best and all the peaks free from cloud. This means a continuous struggle with the weather, and maybe frequent and fruitless ascents, or interruption of the work through bad weather. Versatility and perseverance, reliability and zeal, on the part of experienced mountaineers, must therefore be combined with technical ability if the survey is to be successful.

One more indispensable attribute the operator must possess: a capacity for orienting himself rapidly in a strange country, and for recognizing the most favourable view-points. He must also arrange to distribute his exposures in such a manner that they cover all the valleys, folds, and ramifications of the mountains, without leaving any gaps of importance unphotographed.

Plotting the photographs and preparation of the map.

In the first place plotting from the panoramas is necessary for the triangulation framework. By means of the comparator the angles of the triangles are derived for computation of the triangulation. This is a preliminary process on which I need not enlarge, except to state that if the triangulation has not been carried out rigidly, unusual problems are liable to arise. All points, on which observations and photographs have been taken, must be included in the triangulation, and their co-ordinates and heights computed; in addition, the network should be strengthened by fixing additional points to assist the stereographic plotting, which is to follow.

The preparation of the map may now be started by means of the stereo-autograph, which is an expensive instrument, simple in principle, but complicated in structure and performance. The principle is as follows:

The photographs are examined stereoscopically under a high- powered double microscope. Owing to the relatively long bases the country stands out in marked relief. In the field of view of the double microscope, marks are introduced, which on stereoscopic fusion form a floating mark in the relief view of the country. It is possible to place the floating mark on any particular point, and a system of levers gives the direction of that point.

A factor of great importance is that the floating mark can be set to any given height. Thus the operator, who views the country in relief as well as the floating mark in the double microscope, can move the latter in a horizontal plane, determined by the height to which it is set. By moving the floating mark in this plane in apparent contact with the ground, it follows a contour, and an indicating pencil, coupled with the mechanism, traces on a drawing-table the ground plan of the contour. Thus any required contour which can be traced in the relief view can be drawn with great accuracy. Each base permits a portion of country being mapped, and, if the whole area has been photographed, a contoured plan can be constructed without any gaps. In a similar way other lines and points in the view can be plotted by the autograph, and an accurate and complete map is gradually produced.


Karl Wien

I was myself trained in the technique of the method by Dr. Finsterwalder in Zillertal, and had already used it on the Russo-German Pamir Expedition of 1928. Bauer now put me in charge of the work on the Zemu glacier. The Notgemeinschaft der Deutschen Wissenschaft, with the permission of the President, His Excellency Schmitt-Ott, kindly placed a small photogrammeter with levels (size of plates 9X12 cm.) at our disposal, which being light and easy to handle was most suitable for our work. It had been designed by Dr. Finsterwalder purposely for use by persons without geodetic training and was first used on the Pamir expedition. We took with us 200 plates, made by the firm of Perutz at Munich.

I was ignorant at first to what extent I would be hampered by the steepness and inaccessibility of the mountains and by bad weather. The first of these difficulties caused me to lay out many bases on the hill-sides, regardless of whether the stations would be identifiable from other stations. It was also seldom possible to set up stations on outstanding eminences, which would have been best for the panoramas, or even on spurs so that base-lines could be extended on both sides. One attempt was made at a height of 6,000 m. to take photographs on both sides of a base on the North-east Spur, and this involved us in all kinds of trouble.

We soon found out how much we were at the mercy of the weather. Brenner and I, when we first left Lachen, climbed up from Yaktang to the Yumtso La, and waited there fruitlessly during the 5th, 6th, and 7 th July for a clear day. Twice in the mornings it cleared for a short time, but not for long enough for our work. I had hoped to get a general idea of the whole country from here, but unfortunately was unsuccessful. I had therefore to fix my stations while marching from camp to camp as best I could whenever the mist cleared sufficiently.

After this somewhat fruitless excursion we joined the rest of the party at Camp 3, some kilometres above the snout of the glacier on its northern edge. It was the prearranged starting-point for the work in the region of the lower glacier, just as Camp 6 was to be the centre for the higher work. The only time available for the survey now was while the baggage was being carried to the foot of Kangchenj unga and while the first camps were being erected on its Northeast Spur. Once we began the assault I intended to suspend work, for all our energies must be concentrated on the Spur. I could still complete my photography after the conclusion of the expedition, if fate did not compel us to make too hasty and sudden a retreat. I had therefore arranged with Bauer to complete as much as possible before the assault, though the time and labour were increased out of all proportion to the results.

We soon discovered how bad weather hampered the work, for, although a great amount of preliminary work could be done during fog and rain, no photography was possible as long as there was mist. During August 1929, after snow had fallen all night, it was often clear and cloudless in the morning, which could be utilized for photography. In 1931 it was not so. Of the fourteen days I spent in Camp 3, two days only were clear in the early morning; on the first of these by half-past seven, and on the second by ten o'clock, the clouds from the Zemu valley had banked up so high, that further work was impossible. We remained constantly on the watch for sudden breaks, when it should clear sufficiently for work at a base to be completed. We made all preparations beforehand; but it was not easy to do so, when there was no visibility.

On the 13 th July I photographed from the first base on the south side of the Zemu glacier opposite Camp 3. We-Hermann Schaller and I-had waited for a day in the rain, and made all possible preparations beforehand. We felt instinctively that the morning would be fine, and were ready, as soon as it dawned at 5 a.m., with our instrument set up at the first point. By half-past seven, when the last photograph had been taken from the other end of the base, the peaks were already in cloud, and a few minutes later we were also enveloped.

The next time we had two clear hours was on the 25th July, twelve days later. I now completed the base to the north of the glacier, on the slopes directly above Camp 3, where I had already erected a huge cairn some days before with Schaller. Although I took some further photographs up the glaciers, I finished the whole work, and it was only at 10 a.m., at my last setting, that the mist drew over my cairn and it was lost to view.

Two bases were thus the result of three weeks' work. At this rate I would hardly complete the photography. I went the same day to the 'Green Lake', hoping that another clear morning would follow; but it rained and I could see no further than the moraine. On the 26th I was relieved of charge of Camp 3 and moved up to Camp 6. Here, as the mist lifted, I was able to study the country and so obtained a clear idea of what was required to lay out and complete the work on the upper Zemu glacier. I still had time to descend to the 'Green Lake' and to mark both bases on the east side of the Zemu, without, however, being able to take any photographs.

On the 2nd August I went up to the Adlerhorst, whence we started the assault on the North-east Spur. I intended to observe and photograph at one base on the so-called 'horizontal ridge' (c. 6,050 m.), and, as we had prepared places for two tents, I took the opportunity to have the apparatus brought up by one of the porters who was sent up daily by Aufschnaiter from the Adlerhorst to the Ridge Camp. Unfortunately the ridge is so narrow at this point that it was most difficult to erect the instrument; when set up, it was impossible to approach it from more than one side. On the 7th and 17th August we worked from one of the tents of the Ridge Camp and on the 'Shoulder'; between these dates Schaller's death occurred and we returned to the glacier below. Before and after these dates we went on with the difficult construction of the route up to Camp 8. We might have been able to fix a base-line higher than we did, possibly near Camp 8, where the ground was less difficult and where there was more room to work, but the task of transporting the instrument to such heights depended on local conditions and transport facilities. As it was, the Ridge Camp was the limit to which we could take the instrument, and, when I understood that arrangements for the assault were complete, I sent the instrument down and was relieved to hear that it arrived safely at the Depot at Camp 6.

So it was that before the assault on Kangchenjunga only a portion of the photography was completed, and the question of how I was to finish the work was often the subject of debates in the higher camps on the North-east Spur. When the last detachment retreated to the Adlerhorst on the 22nd September, fine weather set in, and it remained practically clear without the break of as much as a day or even a couple of hours. Our chief difficulties therefore vanished. Nevertheless I was glad of every plate that I had previously exposed, and of every cairn already erected, for our reserves of strength were much reduced by the high camps on the North-east Spur. Even now we had to struggle against time, for fear that the first autumn snowfall might bring the work to a premature end.

From the start I had taken my trusty servant Pemba on all my topographical excursions. He was one of the three porters who accompanied us to the highest camp. He showed the greatest pride in the fact that he alone carried the photographic apparatus, and, indeed, not only carried the instrument but usually the plate-holders and stand as well; and it was this pride that induced him to renounce the privilege granted by Bauer to these three men of carrying no loads after leaving Camp 7. Pemba and Bagde, a young Sherpa, who looked after my personal comfort, remained behind with me on the upper Zemu glacier, when the rest made their way back with the baggage to the Base Camp and Lachen.

On the 26th September I took the photographs on the east side of the upper Zemu glacier opposite Camp 6. They were intended to embrace Kangchenj unga and the south-east side of the North-east Spur, but unfortunately they proved insufficient to include the lower part of the East Ridge of Kangchenjunga and the Zemu gap. I therefore laid out a base-line between a point on the lateral moraine of the glacier and another on the slopes to the west of the glacier.

This was on the day when we finally struck Camp 6, and while we were working on the moraine, our porters clambered down with their heavy loads.

On the 30th September I tried to photograph the ridge running south-east from the 'Sugarloaffrom a base that I had long planned out at a high camp on the lower Twins glacier. Once more we had a race against time, and lost by a few minutes owing to bad weather. We returned to Camp 5, hoping to repeat the attempt in better weather. Here I met Bauer, Hartmann, and Fendt, who left Camp 5 on the 1st October, the last bad day.

Between the 2nd and 10th October I visited a station a day, except on the 8th, when I only made a reconnaissance. I was thus able to complete the topographical work in the estimated time. I first considered establishing a high base on the upper ridge of the 'Sugarloaf', but I would have required a whole day for the observa- tions. Though the weather was now perfect, clouds coming up the Zemu valley obstructed the view and hindered work in the afternoon. Moreover, I could make three bases sufficiently high on the 'Sugar- loaf, opposite Camp 5, and on the rocky crags above the Green Lake, which should be sufficient for my purpose. On the 2nd, therefore, we crossed the Zemu glacier from Camp 5, a tiring performance, and took photographs from 'under the Red Tower towards the west, which I had reconnoitred with Pemba as long ago as in July. On the 3rd I climbed the slope by Camp 5 to make further observations from 'Sugarloaf. The way was long and difficult, and we only arrived at eleven o'clock after a start at 6.30 a.m. Unfortunately the peaks had already clouded over, so that on the 4th I had to go up again for the third time, completing the remaining observations in perfect weather by nine o5 clock.

On the 5th October we took the camp down to the Green Lake. On the way Pemba and I, just before the junction of the Tent Peak glacier, climbed about 200 metres up the lower slopes of the 'Sugar- loaf5, and established one more base facing the Simvu Saddle. On the 6th we were on the rocks above the Green Lake, whence we had a magnificent view eastwards to the snout of the glacier, as well as westwards towards Tent Peak and Tent Peak glacier. On this day I exposed over twenty plates.

On the 7th October we were on the south side of the glacier again, where under unfavourable conditions and in a bad state of health I photographed from a rather low base towards the northern mountains. I had hoped to get far enough up the Tent Peak glacier to have a clear view of the Nepal gap, but a reconnaissance on the 8th showed that it was impracticable. The following day, after a snowfall at night, we descended through a wintry landscape to Camp 3. On the way we took photographs of points that we had marked previously in July towards Siniolchu. On the next day, the 19th October, I went down with the remaining plates to the lake below Camp 3. It was an absolutely perfect day. I climbed sufficiently high for the whole range from Lamgebo up to Kangchenjunga and the Tent Peak to be visible. We had now established ten groups of bases, five on each side of the upper and lower Zemu glacier, one on the flat top of the North-east Spur; this was to be the last, and was to include the snout of the glacier.

The weather was so settled that I allowed myself plenty of time. When the last plate was exposed the survey was finished and we left Camp 3 and the Zemu valley. I marched back by the Yumtso La and Mangen to Darjeeling with a German traveller who had come up from there to the Zemu glacier. Now we had to await the development of the plates, which could not be done till we reached home. I often wondered on the way home how the plates were keeping in the cold-storage room of our steamer, and how the results would turn out. We had the great good fortune that Dr. Finsterwalder undertook the work, and we offer him our heartfelt thanks that he brought everything to a satisfactory conclusion in spite of all difficulties.


Richard Finsterwalder

The Framework.

The photographs brought back by Wien, 195 in all, were without exception successful. Sixty-three of them form the basis of the triangulation, while the rest were to be used for plotting with the autograph.[5] Those intended for the triangulation were taken from thirteen stations. At four stations complete circular panoramas were taken, with eight exposures at each, at three stations there were incomplete panoramas, and at the others single photographs were taken.

For the autograph plotting there were twenty-one base-lines with fifty-one pairs of photographs. All data relating to the panoramas and bases were explained clearly with illustrative sketches in a field- book, a very necessary precaution. Accurate angular measurements were not made for the main triangulation, but compass readings were taken, so that the points could be determined. The material collected by Wien was amplified by a number of valuable points, whose co-ordinates and heights were taken from the Survey of India Triangulation Pamphlet 78A, p. no. These points-I refer to the high peaks towering above the range-were fixed from stations of the primary triangulation to the south. Their positions are given in geographical co-ordinates, mostly to the nearest second of latitude and longitude, but sometimes to tenths of a second.

Computing the Network.

Considerable difficulties had to be overcome and it was long doubtful whether much would result. Apparently owing to a flaw in t he photogrammeter compass, the readings on which the computation of the network and lengths of base-lines were dependent showed errors up to 10° and so could not be utilized. A comprehensive triangulation, which might have given approximate positions for many points, could not therefore be undertaken.

An attempt was then made to find the images of the stations on the plates and to measure the angles between the rays joining them. By this means we might be able to build up a subsidiary network which could be fitted to the points of the Indian Survey triangulation. But here too there were great difficulties; for Wien, exhausted at the end of the assault on Kangchenjunga, had been unable to place his stations on prominent and lofty points. Some were down the slopes of ridges, and, although marked by cairns, were not readily identifiable on the plates taken from a distance of several kilometres. From the few and rather uncertain angles obtained by this means it proved impossible to build up a net.

A final way out of our difficulties was to fall back on the points of the Indian triangulation, though there were some important considerations to take into account. The co-ordinates of all, except those of Kangchenjunga itself, are given only to whole seconds of latitude and longitude, which means a possible error of 30 m. in both co-ordinates. The errors are not given, but are at least 30 m., and had to be assumed even greater, for we were dealing with points observed from great distances and from several different stations. There was also the uncertainty of identification, since only a few of them have well-defined highest points, to which the co-ordinates refer. A careful examination of the photographs gave us clear identifications for three points fixed by the Survey, Kangchenjunga, Siniolchu, and Lamgebo (east of Siniolchu, towards the lower end of the Zemu valley); Nepal Peak was not quite so reliable.1 The sharpness of these peaks makes it probable that they were well identified by the Survey of India from every direction, and are therefore well fixed.

As Wien had taken no photographs on any of the Indian Triangulation stations, only resections were possible for the fixing of the various panorama stations. For a resection, however, rays to three well-placed and well-fixed points are necessary, and at least another ray to a well-known point as a check. There were data for the absolute fixing of only two points by this method; the panoramas from the remainder enabled three more stations to be recognized; for the rest the data were still less favourable.

Difficulties were finally overcome by utilizing the rays to the Indian points as well as those found from base to base. After careful study of the plates we proceeded step by step, fixing new points throughout the network by forward intersection from each newly- fixed station, and these new points gave the necessary data for resecting the missing stations.

1 The number of summits in the neighbourhood of Simvu and Little Siniolchu made identification unreliable. The position of 'The Twins' appeared wrong and that of Simvu unreliable.

[ The original text here gives each step of the resection and intersection of the main network. Owing to lack of space, these details have been omitted. The work of computation was done by Mr. Karl Reicheneder of Munich. It comprised measurements from sixty-three plates, and took six weeks. A chart of the triangulation net appears in Um den Kantsch! p. 122, and a complete list of co-ordinates and heights (with probable errors) of stations of the network, and of base stations, is given on pp. 129, 130.-Ed.]

Computation of the Photogrammetric Stations.

As far as possible the stereo-photogrammetric stations had been arranged to coincide with the panorama stations of the triangulation network. Thus there was little more to be computed at the stereo- pi 10 togrammetric stations than the length and other details of the base. Owing to the compass error already mentioned, there was some uncertainty which had to be remembered when plotting on the Autograph.

Determination of Heights.

The angles for height were measured from the plates at the same time as the horizontal angles. Differences of height between stations were thus obtained. They are based on the Survey of India height lor Kangchenj unga,[6] 28,146 feet (=8,579 m.). Probable errors in height were small, and for all points less than 10 m.

The Plotting on the Stereo-Autograph.

All this was work preparatory to the actual plotting on the Zeiss Stereo-Autograph, undertaken by Hans Biersack, whose long experience of such work gave him great dexterity. There were again some difficulties, owing to the inexactness of some of the base determinations, the short lengths of others, and the consequent lack of stereoscopic effect for distant slopes. On the other hand, Wien had been successful in photographing practically the whole of the lower Zemu glacier, almost without a gap.2
The material was exploited to the utmost limit, and work was extended to country which was really too far from some of the bases, in order to confirm contour details. This happened on the southern slopes of Tent and Nepal Peaks; and for this reason the contours here have been intentionally more generalized. Again, occasional uncertainty was traceable to the fact, already mentioned, that coordinates of initial points were only correct to about 30 m.; such doubt was eliminated when the plottings from the various bases were joined up. The adjustments were small but unavoidable.

The plotting of fifty-one pairs of plates, belonging to twenty-one different bases, on a scale of 1 : 25,000, took five weeks. It was done at the Technical High School at Munich. The resulting contours express the rugged form of the Kangchenjunga massif with great fidelity, though the contour interval, which could not be less than 50 m., owing to the steepness of the ground, sets a limit to the representation of some of the minor features which exist between them. For this reason the contours on the flat trunk of the Zemu glacier have been traced at the closer interval of 25 m.

Further Elaboration of the Map.

Additional detail to that obtained from the autograph was added. The larger features have been shown up by hill-shading, drawn first on tracing paper and photographed through a screen and surprinted as a flat tint. Such a method is somewhat primitive, and tends to reduce the ruggedness of the features, but the superior processes, relief and intaglio printing and collotype, are too costly for large maps. The simple method of hill-shading also permitted a generalized idea of the topography to be carried to the borders of the sheet, as, for instance, in the area of the Passanram glacier, south of Siniolchu and Simvu, which was sketched from amateur photographs taken by Allwein. For the western slopes of the Kangchenjunga- Tent Peak ridge, the rough representation is taken from the Kangchenjunga map by Marcel Kurz.

The Scale of the Map, 1 :33,333 (3 cm. = 1 km.).

The scale of the map has been the subject of discussion in cartographic circles. The deciding factors were these: The larger scale of 1 : 25,000 would have given a map too large to handle and too expensive to produce, while the smaller scale of 1 : 50,000 would hardly have allowed sufficient detail to be shown. The scale of 1 :33,333 gives us a handy map, from which measurements with a centimetre rule are easily taken, a kilometre on the ground exactly corresponding to 3 cm. on the map. The chief disadvantage to this scale is that comparison with existing maps is not easy.

[1] By the courtesy of Herr Paul Bauer and of the Deutscher und Oesterreichischer Alpenverein the map of the Zemu glacier is issued with this volume of the Himalayan Journal. A photographic reproduction of a preliminary map of Nanga Parbat, produced by the same stereo-photogrammetric method, accompanies the paper on the 'Scientific Results of the Nanga Parbat Expedition, 1934'.

This paper appeared in German in Paul Bauer's Um denKantsch! (pp. 107-30), which was reviewed in Himalayan Journal, vol. v, 1933, pp. 143-6- I have edited Colonel Thompson's literal translation into freer English, in order to save space and avoid repetition.-Ed.

[2] See also Geographical Journal,

1902, vol. xx, pp. 13, 14.

[3] Dr. Kellas's map, Alpine Journal, May 1912, should also be mentioned. It introduces certain improvements into Garwood's map, but otherwise shows no progress as regards cartography.

[4] Supplementary map to G. O. Dyhrenfurth's Himalaya.

[5] The directions of the photographs are shown on the map by small arrows. The photogrammeter was built to my design for the Alai-Pamir Expedition of 1928 by O. Steinheil & Sons, and lent by the Notgemeinschaft der Deutschen Wissenschaft.

[6] See also In Kampf um den Himalaya (Paul Bauer), Munich, 1931, p. 153.

For the north side of the ridge between the Nepal gap and 'SugarloafHerr I Mrich Wieland kindly placed some photographs at our disposal. We wish here to u knowledge our thanks for these.-P. Bauer.