My observations for temperature and wet-bulb being for the most part desultory, taken at different dates, and under very different conditions of exposure, etc., it is obvious that those at one station are hardly, if at all, comparative with those of another, and I have therefore selected only such as were taken at the same date and hour with others taken at the Calcutta Observatory, or as can easily be reduced; which thus afford a standard (however defective in many respects) for a comparison. I need hardly remind my reader that the vapour-charged wind of Sikkim is the southerly one, which blows over Calcutta; that in its passage northwards to Sikkim in the summer months, it traverses the heated plains at the foot of the Himalaya, and ascending that range, it discharges the greater part of its moisture (120 to 140 inches annually) over the outer Himalayan ranges, at elevations of 4000 to 8000 feet. The cooling effect of the uniform covering of forest on the Sikkim ranges is particularly favourable to this deposition, but the slope of the mountains being gradual, the ascending currents are not arrested and cooled so suddenly as in the Khasia mountains, where the discharge is consequently much greater. The heating of the atmosphere, too, over the dry plains at the foot of the outer range, increases farther its capacity for the retention of vapour, and also tends to render the rain-fall less sudden and violent than on the Khasia, where the south wind blows over the cool expanse of the Jheels. It will be seen from the following observations, that in Sikkim the relative humidity of the atmosphere remains pretty constantly very high in the summer months, and at all elevations, except in the rearward valleys; and even there a humid atmosphere prevails up to 14,000 feet, everywhere within the influence of the snowy mountains. The uniformly high temperature which prevails throughout the summer, even at elevations of 17,000 and 18,000 feet, is no doubt proximately due to

[ 423 ]

the evolution of heat during the condensation of these vapours. It will be seen by the pages of my journal, that continued sunshine, and the consequent heating of the soil, is almost unknown during the summer, at any elevation on the outer or southward ranges of Dorjiling: but the sunk thermometer proves that in advancing northward into the heart of the mountains and ascending, the sun's effect is increased, the temperature of the earth becoming in summer considerably higher than that of the air. With regard to the observations themselves, they may be depended upon as comparable with those of Calcutta, the instruments having been carefully compared, and the cases of interpolation being few. The number of observations taken at each station is recorded in a separate column; where only one is thus recorded, it is not to be regarded as a single reading, but the mean, of several taken during an hour or longer period. I have rejected all solitary observations, even when accompanied by others at Calcutta; and sundry that were, for obvious reasons, likely to mislead. Where many observations were taken at one place, I have divided them into sets, corresponding to the hours at which alone the Calcutta temperature and wet-bulb thermometer are recorded,* in order that meteorologists may apply them to the solution of other questions relating to the distribution of heat and moisture. The Dorjiling observations, and those in the immediate neighbourhood of that station, appeared to me sufficiently numerous to render it worth while classing them in months, and keeping them in a series by themselves. The tensions of vapour are worked from the wet-bulb readings by Apjohn's formula and tables, corrected for the height of the barometer at the time. The observations, except where otherwise noted, are taken by myself.

* Sunrise; 9.50 a.m.; noon; 2.40 p.m.; 4 p.m., and sunset.

[ 424 ]

SERIES I. Observations made at or near Dorjiling.

JANUARY, 1849

* Observations to which the asterisk is affixed were taken by Mr. Muller.

JANUARY, 1850

[ 425 ]

FEBRUARY, 1850

* Observations to which the asterisk is affixed were taken by Mr. Muller.

MARCH, 1850

[ 426 ]

APRIL

* Observations to which the asterisk is affixed were taken by Mr. Muller.

MAY

JUNE

[ 427 ]

JULY, 1848

* Observations to which the asterisk is affixed were taken by Mr. Muller.

AUGUST, 1848

* Observations to which the asterisk is affixed were taken by Mr. Muller.

SEPTEMBER, 1848

* Observations to which the asterisk is affixed were taken by Mr. Muller.

[ 428 ]

OCTOBER, 1848

NOVEMBER AND DECEMBER, 1848

* Observations to which the asterisk is affixed were taken by Mr. Muller.

[ 429 ]

Comparison of Dorjiling and Calcutta.

It is hence evident, from nearly 1000 comparative observations, that the atmosphere is relatively more humid at Dorjiling than at Calcutta, throughout the year. As the southerly current, to which alone is due all the moisture of Sikkim, traverses 200 miles of land, and discharges from sixty to eighty inches of rain before arriving at Dorjiling, it follows that the whole atmospheric column is relatively drier over the Himalaya than over Calcutta; that the absolute amount of vapour, in short, is less than it would otherwise be at the elevation of Dorjiling, though the relative humidity is so great. A glance at the table at the end of this section appears to confirm this; for it is there shown that, at the base of the Himalaya, at an elevation of only 250 feet higher than Calcutta, the absolute amount of vapour is less, and of relative humidity greater, than at Calcutta.

[ 430 ]

SERIES II. Observations at various Stations and Elevations in the Himalaya of East Nepal and Sikkim.

ELEVATION 735 TO 2000 FEET

ELEVATION 2000 TO 3000 FEET

[ 431 ]

ELEVATION 3000 TO 4000 FEET

ELEVATION 4000 TO 5000 FEET

[ 432 ]

ELEVATION 5000 TO 6000 FEET

[ 433 ]

ELEVATION 6000 TO 7000 FEET

ELEVATION 7000 TO 8000 FEET

[ 434 ]

ELEVATION 8000 TO 9000 FEET

ELEVATION 9000 TO 10,000 FEET

[ 435 ]

ELEVATION 10,000 TO 11,000 FEET

ELEVATION 11,000 TO 12,000 FEET

[ 436 ]

ELEVATION 12,000 TO 13,000 FEET

ELEVATION 13,000 TO 14,000 FEET

[ 437 ]

ELEVATION 15,000 TO 16,000 FEET

ELEVATION 16,000 TO 17,000 FEET

[ 438 ]

ELEVATION 17,000 TO 18,500 FEET

SUMMARY

Considering how desultory the observations in Sikkim are, and how much affected by local circumstances, the above results must be considered highly satisfactory: they prove that the relative humidity of the atmospheric column remains pretty constant throughout all elevations, except when these are in a Tibetan climate; and when above 18,000 feet, elevations which I attained in fine weather only. Up to 12,000 feet this constant humidity is very marked; the observations made at greater elevations

[ 439 ]

were almost invariably to the north, or leeward of the great snowy peaks, and consequently in a drier climate; and there it will be seen that these proportions are occasionally inverted; and in Tibet itself a degree of relative dryness is encountered, such as is never equalled on the plains of Eastern Bengal or the Gangetic delta. Whether an isolated peak rising near Calcutta, to the elevation of 19,000 feet, would present similar results to the above, is not proven by these observations, but as the relative humidity is the same at all elevations on the outermost ranges of Sikkim, which attain 10,000 feet, and as these rise from the plains like steep islands out of the ocean, it may be presumed that the effects of elevation would be the same in both cases.

The first effect of this humid wind is to clothe Sikkim with forests, that make it moister still; and however difficult it is to separate cause from effect in such cases as those of the reciprocal action of humidity on vegetation, and vegetation on humidity, it is necessary for the observer to consider the one as the effect of the other. There is no doubt that but for the humidity of the region, the Sikkim Himalaya would not present the uniform clothing of forest that it does; and, on the other hand, that but for this vegetation, the relative humidity would not be so great.*

The great amount of relative humidity registered at 6000 to 8000

* Balloon ascents and observations on small mountainous islands, therefore, offer the best means of solving such questions: of these, the results of ballooning, under Mr. Welsh's intrepid and skilful pioneering (see Phil. Trans. for 1853), have proved most satisfactory; though, from the time for observation being short, and from the interference of belts of vapour, some anomalies have not been eliminated. Islands again are still more exposed to local influences, which may be easily eliminated in a long series of observations. I think that were two islands, as different in their physical characters as St. Helena and Ascension, selected for comparative observations, at various elevations, the laws that regulate the distribution of humidity in the upper regions might be deduced without difficulty. They are advantageous sites, from differing remarkably in their humidity. Owing partly to the indestructible nature of its component rock (a glassy basalt), the lower parts of Ascension have never yielded to the corroding effects of the moist sea air which surrounds it; which has decomposed the upper part into a deep bed of clay. Hence Ascension does not support a native tree, or even shrub, two feet high. St. Helena, on the other hand, which can hardly be considered more favourably situated for humidity, was clothed with a redundant vegetation when discovered, and trees and tree-ferns (types of humidity) still spread over its loftiest summits. Here the humidity, vegetation, and mineral and mechanical composition reciprocate their influences.

[ 440 ]

feet, arises from most of the observations having been made on the outer range, where the atmosphere is surcharged. The majority of those at 10,000 to 12,000 feet, which also give a disproportionate amount of humidity, were registered at the Zemu and Thlonok rivers, where the narrowness of the valleys, the proximity of great snowy peaks, and the rank luxuriance of the vegetation, all favour a humid atmosphere.

I would have added the relative rain-fall to the above, but this is so very local a phenomenon, and my observations were so repeatedly deranged by having to camp in forests, and by local obstacles of all kinds, that I have suppressed them; their general results I have given in Appendix F.

I here add a few observations, taken on the plains at the foot of the Sikkim Himalaya during the spring months.

Comparison between Temperature and Humidity of the Sikkim Terai and Calcutta, in March and April, 1849.

From the above, it appears that during the spring months, and before the rains commence, the belt of sandy and grassy land along the Himalaya, though only 3·5° north of Calcutta, is at least 6° or 7° colder, and always more humid relatively, though there is absolutely less moisture suspended in the air. After the rains commence; I believe that this is in a great measure inverted, the plains becoming

[ 441 ]

excessively heated, and the temperature being higher than at Calcutta. This indeed follows from the well known fact that the summer heat increases greatly in advancing north-west from the Bay of Bengal to the trans-Sutledge regions; it is admirably expressed in the maps of Dove's great work "On the Distribution of Heat on the Surface of the Globe."

These observations were taken by burying a brass tube two feet six inches to three feet deep, in exposed soil, and sinking in it, by a string or tied to a slip of wood, a thermometer whose bulb was well padded with wool: this, after a few hours' rest, indicates the temperature of the soil. Such a tube and thermometer I usually caused to be sunk wherever I halted, if even for one night, except during the height of the rains, which are so heavy that they communicate to the earth a temperature considerably above that of the air.

The results proved that the temperature of the soil at Dorjiling varies with that of the month, from 46° to 62·2°, but is hardly affected by the diurnal variation, except in extreme cases. In summer, throughout the rains, May to October, the temperature is that of the month, which is imparted by the rain to the depth of eleven feet during heavy continued falls (of six to twelve inches a day), on which occasions I have seen the buried thermometer indicating a temperature above the mean of the month. Again, in the winter months, December and January, it stands 5° above the monthly mean; in November and February 4° to 5°; in March a few degrees below the mean temperature of the month, and in October above it; April and May being sunny, it stands above their mean; June to September a little below the mean temperature of each respectively.

The temperature of the soil is affected by:—1. The exposure of the surface; 2. The nature of the soil; 3. Its permeability by rain, and the presence of underground springs; 4. The sun's declination; 5. The elevation above the sea, and consequently the heating power of the sun's rays: and 6, The amount of cloud and sunshine.

The appended observations, though taken at sixty-seven places, are

[ 442 ]

far from being sufficient to supply data for the exact estimation of the effects of the sun on the soil at any elevation or locality; they, however, indicate with tolerable certainty the main features of this phenomenon, and these are in entire conformity with more ample series obtained elsewhere. The result, which at first sight appears the most anomalous, is, that the mean temperature of the soil, at two or three feet depth, is almost throughout the year in India above that of the surrounding atmosphere. This has been also ascertained to be the case in England by several observers, and the carefully conducted observations of Mr. Robert Thompson at the Horticultural Society's Gardens at Chiswick, show that the temperature of the soil at that place is, on the mean of six years, at the depth of one foot, 1° above that of the air, and at two feet 1·5°. During the winter months the soil is considerably (l° to 3°) warmer than the air, and during summer the soil is a fraction of a degree cooler than the air.

In India, the sun's declination being greater, these effects are much exaggerated, the soil on the plains being in winter sometimes 9° hotter than the air; and at considerable elevations in the Himalaya very much more than that; in summer also, the temperature of the soil seldom falls below that of the air, except where copious rain-falls communicate a low temperature, or where forests interfere with the sun's rays.

At considerable elevations these effects are so greatly increased, that it is extremely probable that at certain localities the mean temperature of the soil may be even 10° warmer than that of the air; thus, at Jongri, elevation 13,194 feet, the soil in January was 34·5°, or 19·2° above the mean temperature of the month, immediately before the ground became covered with snow for the remainder of the winter; during the three succeeding months, therefore, the temperature of the soil probably does not fall below that of the snow, whilst the mean temperature of the air in January may be estimated at about 20°, February 22°, March 30°, and April 35°. If, again, we assume the temperature of the soil of Jongri to be that of other Sikkim localities between 10,000 and 14,000 feet, we may assume the soil to be warmer by 10° in July (see Tungu observations), by 8° or 9° in September (see Yeumtong); by l0° in October (see Tungu); and by 7° to l0° in November (see Wallanchoon and Nanki). These temperatures,

[ 443 ]

however, vary extremely according to exposure and amount of sunshine; and I should expect that the greatest differences would be found in the sunny climate of Tibet, where the sun's heat is most powerful. Were nocturnal or terrestrial radiation as constant and powerful as solar, the effects of the latter would be neutralised; but such is not the case at any elevation in Sikkim.

This accumulated heat in the upper strata of soil must have a very powerful effect upon vegetation, preventing the delicate rootlets of shrubs from becoming frozen, and preserving vitality in the more fleshy, roots, such as those of the large rhubarbs and small orchids, whose spongy cellular tissues would no doubt be ruptured by severe frosts. To the burrowing rodents, the hares, marmots, and rats, which abound at 15,000 to 17,000 feet in Tibet, this phenomenon is even more conspicuously important; for were the soil in winter to acquire the mean temperature of the air, it would take very long to heat after the melting of the snow, and indeed the latter phenomenon would be greatly retarded. The rapid development of vegetation after the disappearance of the snow, is no doubt also proximately due to the heat of the soil, quite as much as to the increased strength of the sun's direct rays in lofty regions.

I have given in the column following that containing the temperature of the sunk thermometer, first the extreme temperatures of the air recorded during the time the instrument was sunk; and in the next following, the mean temperature of the air during the same period, so far as I could ascertain it from my own observations.

SERIES I. Soane Valley

[ 444 ]

SERIES II. Himalaya of East Nepal and Sikkim

* Sheltered by trees, ground spongy and wet.

[ 445 ]

SERIES III. Plains of Bengal

SERIES IV. Khasia Mountains

* Hole full of rain-water.     † Soil, a moist sand.     § Dry sand.

[ 446 ]

SERIES V. Jheels, Gangetic Delta, and Chittagong

* Shaded by trees.     † Observations at the Mint, etc., by Mr. Muller.
§ Observations for temperature of air, taken at the Observatory.

I have selected as many of my observations for temperature of the sir as appeared to be trustworthy, and which, also, were taken contemporaneously with others at Calcutta, and I have compared them with the Calcutta observations, in order to find the ratio of decrement of heat to an increase of elevation. The results of several sets of observations are grouped together, but show so great an amount of discrepancy, that it is evident that a long series of months and the selection of several stations are necessary in a mountain country to arrive at any accurate results. Even at the stations where the most numerous and the most trustworthy observations were recorded, the results of different months differ extremely; and with regard to the other stations, where few observations were taken, each one is affected differently from another at the same level with it, by the presence or proximity of forest, by exposure to the east or west, to ascending or

[ 447 ]

descending currents in the valleys, and to cloud or sunshine. Other and still more important modifying influences are to be traced to the monthly variations in the amount of humidity in the air and the strength of its currents, to radiation, and to the evolution of heat which accompanies condensation raising the temperature of elevated regions during the rainy season. The proximity of large masses of snow has not the influence I should have expected in lowering the temperature of the surrounding atmosphere, partly no doubt because of the more rapid condensation of vapours which it effects, and partly because of the free circulation of the currents around it. The difference between the temperatures of adjacent grassy and naked or rocky spots, on the other hand, is very great indeed, the former soon becoming powerfully heated in lofty regions where the sun's rays pass through a rarefied atmosphere, and the rocks especially radiating much of the heat thus accumulated, for long after sunset. In various parts of my journals I have alluded to other disturbing causes, which being all more or leas familiar to meteorologists, I need not recapitulate here. Their combined effects raise all the summer temperatures above what they should theoretically be.

In taking Calcutta as a standard of comparison, I have been guided by two circumstances; first, the necessity of selecting a spot where observations were regularly and accurately made; and secondly, the being able to satisfy myself by a comparison of my instruments that the results should be so far strictly comparable.

I have allowed 1° Fahr. for every degree in latitude intervening between Sikkim and Calcutta, as the probable ratio of diminution of temperature. So far as my observations made in east Bengal and in various parts of the Gangetic delta afford a means of solving this question, this is a near approximation to the truth. The spring observations however which I have made at the foot of the Sikkim Himalaya would indicate a much more rapid decrement; the mean temperature of Titalya and other parts of the plains south of the forests, between March and May being certainly 6°–9° lower than Calcutta: this period however is marked by north-west and north-east winds, and by a strong haze which prevents the sun's rays from impinging on the soil with any effect. During the southerly

[ 448 ]

winds, the same region is probably hotter than Calcutta, there being but scanty vegetation, and the rain-fall being moderate.

In the following observations solitary readings are always rejected.

Observations taken during the rainy season of 1848, at Mr. Hodgson's (Jillapahar, Dorjiling) alt. 7,430 feet, exposure free to the north east and west, the slopes all round covered with heavy timber; much mist hence hangs over the station. The mean temperatures of the month at Jillapahar are deduced from horary observations, and those of Calcutta from the mean of the daily maximum and minimum.

[ 449 ]

These, it will be seen, give a result which approximates to that of the sets I and II. Being deduced from observations at different exposures, the effects of these may be supposed to be eliminated. It is to be observed that the probable results of the addition of November and December's observations, would be balanced by those of May and June, which are hot moist months.

[ 450 ]

The total number of comparative observations taken during that journey, amounted to 563, and the mean equivalent was 1°=303 feet, but I rejected many of the observations that were obviously unworthy of confidence.

The great elevation of the temperature in the lowest elevations is accounted for by the heating of the valleys wherein these observations were taken, and especially of the rocks on their floors. The increase with the elevation, of the three succeeding sets, arises from the fact that the loftier regions are far within the mountain region, and are less forest clad and more sunny than the outer Himalaya.

A considerable number of observations were taken during this journey at night, when none are recorded at Calcutta, but which are comparable with contemporaneous observations taken by Mr. Muller at Dorjiling. These being all taken during the three most rainy months, when the temperature varies but very little during the whole twenty-four hours, I expected satisfactory results, but they proved very irregular and anomalous.

The means were—
           At 21 stations of greater elevation than Dorjiling 1°=348 ft.
           At 17 stations lower in elevation 1°=447 ft.

[ 451 ]

The above results would seem to indicate that up to an elevation of 7,500 feet, the temperature diminishes rather more than 1° Fahr. for every 300 feet of ascent or thereabouts; that this decrement is much leas in the summer than in the winter months; and I may add that it is less by day than by night. There is much discrepancy between the results obtained at greater or less elevations than 7000 feet; but a careful study of these, which I have arranged in every possible way, leads me to the conclusion that the proportion map be roughly indicated thus:—

[ 452 ]

The equivalent thus deduced is far greater than that brought out by the Sikkim observations. It indicates a considerably higher temperature of the atmosphere, and is probably attributable to the evolution of heat during extraordinary rain-fall, and to the formation of the surface, which is a very undulating table-land, and everywhere traversed by broad deep valleys, with very steep, often precipitous flanks; these get heated by the powerful sun, and from them, powerful currents ascend. The scanty covering of herbage too over a great amount of the surface, and the consequent radiation of heat from the earth, must have a sensible influence on the mean temperature of the summer months.

Appendix A | Appendix B | Appendix C | Appendix D | Appendix E | Appendix F |
Appendix G | Appendix H | Appendix I | Appendix J | Appendix K | Appendix L |