J Sg 640” 6.

= -t: 73!

. , 4 }

ise toeceor .

4 ve!

mm ee eee es eee

one

+ «1%

——_-

_—— .

~~

-

“—=— ~~ wwe eee eS ee

or * | 7 ¢¢ > Poe #- 4 ‘ew 6 4 6: a a €- + 7, * ag % 4

ea :

ae a

. ahh tle 3”

"+

* * om

. *, “een i” @

~ *

ob e

oe

BH

be be

JOURNAL

OF THE

ASIATIC SOCIETY OF BENGAL.

—QG-—

WOE. ENE. PART II. (NATURAL HisToRY, &c.)

(Nos. I. to V.—1887.)

EDITED BY

JHE NaTURAL ftisTORY PECRETARY.

~~. sense 00 eee eee sc 0—_ Oo ees

‘¢Tt will flourish, if naturalists, chemists, antiquaries, philologers, and men of science in different parts of Asza will commit their observations to writing, and send them to the Asiatic Society at Calcutta. It will languish, if such communications shall be long intermitted ; and it will die away, if they shall entirely cease.” SIR WM. JONES.

PDOs ess ee eee OOO eer

CALCUTTA: AO!

AND PUBLISHED BY THE

§ KX PRINTED BY G. H. ROUSE, AT THE BAPTIST MISSION PRESS, -

ASIATIC SOCIETY, 57, PARK STREBT. ut

1888, RS

LIST OF CONTRIBUTORS.

—@—

Atkinson, HE. T.;—WNotes on Indian Rhynchota: Heteroptera MM oa thst id 2 dala bs chime cleaned disse nae cee hvsea tes nem es mans ——— ;—Notes on Indian Rhynchota : snibinhii NNR 5 taceicarsantiin ss k cals saoimctule e's Sous namo ir mamak ba oa lee es Seas Barctay, A.;—A apes pe of the Peadinns occurring in the Becbhairhoad : Simla (Western Himalaya). (Plates XII.— Buanvorn, ars BR ate He Dei of adios Bacaite on the Faiegall.. (WV ith. a Wood-cnt),.- sos rancssvesnpsssccsesvsccesepneeses Carpenter, AtrreD ;—Natural History Notes from H. M.’s tees Marine Survey Steamer Investigator,’ Commander ALFRED Carpenter, BR. N., Commanding. No. 7. The Mean Tempera- twre of the Deep Waters of the Bay of Bengal. (Plate X.),... Exuson, Samuet R. ;—On the changes observed in the Density of the Surface Sea-water, coincident with, and due to Aerial Distur- bances, and consequent Alteration of Baric Pressure over adja- cent Sea Areas: and onthe Usefulness of a more exact measure- ment of the Specific Gravity of Sea-water: more specially with Reference to the Waters near and about the Hooghly River Pilot Station. (Diagram—Plate IX.), wicscerccccscescveccesse sas Footr, R. Bruce ;—Notes on some recent Neolithic and Palcoli- thic Finds in South India. (With a Map—Plate XI.), ...... Gitts, G. M. ;—Natural History Notes from H. M.’s Indian Marine Survey Steamer Investigator,’ Commander ALFRED CARPENTER, R. N., Commanding. No. 6. On Six new Amphipods from the Bay of Bengal. (Plates ITT.—VIIL.),........cccecsccscsecces Jones, E. J.;—Natural History Notes from H. M.’s Indian Marine Survey Steamer Investigator,’ Commander ALFRED CARPENTER, &. N., Commanding. No. 5. On some Nodular Stones obtained by trawling off Colombo in 675 Fathoms of Water. (Plate II), Kine, Grorce ;—A second series of New Species of Ficus from New Guined,... 1 cee Pansantie ajar Sinaia Saleen ON some 5 Mets ree of Pied Taain Sumatra, . ;—On the Species of Loranthus indigenous to Perse

Page 22

145

300

230

15

259

212

209

61 65 89

iv List of Contributors.

Movxnorapnyay, AsutosH ;—On the Differential rea a of a Trajectory. (With a Wo0d-cut), ....:..ccccresovavvcresceceravecan ses —— —— ;—On ee Differ nite “aviation to

all Conics,.... a cree , et ot ——_——_—___— ae Maibiro on ; Plone Bait yi Gamma (With three Wood-cuts), ......... Scutty, J.;—On the Mammals and Bide eee an ‘Capa C. E. Yats, C.S.1., of the Afghan Boundary Commission,...... ;—On the Effects produced by Small Quantities of Bts- muth on the Ductility of Silver,. Piven psnWeeaeasete estes ;—On the Chiroptera of ees, A is\siusiniteliclestslehi-teslaleniae ats Smon, H. ;—H’ tude sur les Arachnides de 1 Aste méridionale faisant partie des collections de V Indian Musewm (Calcutta), .......s000 ;—LH’ tude sur les Arachnides del Asie méridionale faisant partie des collections deV Indian Musewm (Calcutta), .......00008 Woop-Mason, J. ;—Natural History Notes from H. M.’s Indian Marine Survey Steamer Investigator,’ Commander ALFRED CarPENntTER, f., N. Commanding. No. 4. Description of a new Species of Crustacea belonging to the pie aead Family Hammiag.’ Chlnte Dy). .02. sistas cesssvere nes ;—Natural History Noise From H. Mw s tale Marine Survey Steamer Investigator,’ Commander ALFRED Carpentrer, &. N., Commanding. No. 8. Description of a new Species of the Brachyurous Genus Lyreidus from the Depths of the Andaman Sed, sessreeceerere

Peeeeeesevese are ake eos ateateser one 188

eae eee eee

Page 116 134 288

68

121 233

101

282

206

376

Dates of issue of the different numbers of the Journal, Part I7,-1887.

No. I.—Containing pp. 1—120, with Plate IX, was issued on July 23rd, 1887.

No. II.—Containing pp. 121—232, with Plates I, II, III, IV, V, VI, VII, & VIII, was issued on November 2nd, 1887.

No. III.—Containing pp. 233—349, with Plates X & XI, was issued on January 30th, 1888.

No. IV.—Containing pp. 349—376, with Plates XII, XIII, XIV, & XV, was issued on March 18th, 1888.

No. V.—Containing Title-page, Index, &c., to the Volume.

LIST OF PLATES.

I. Lyreidus channert. II. Barium nodules from 675 fathoms off Colombo. III. Phronima bucephala and Phronimella hippocephala. IV. Rhabdosoma investigatoris. V. Amphipronoé longicornuta. VI. Lestrigonus bengalensis. VII. VIII. LHurystheus hirsutus. IX. Diagram for Corrections for Temperature of Fresh Water. X. Chart shewing Mean Temperature of the Deep Waters of the Bay of Bengal. XI. Map of South India shewing the chief localities at which Neolli- thic and Paleolithic Remains have been found.

RET. XIII. { Uredines occurring in the Neighbourhood of Simla (Western XIV. Himalaya).

i n

“f), "a iy ee e e a

California Academy of Sciences

Presented a eee

ae

April ra ; 1908.

¥ 7 » a ft . tor «p< eiezus f } 7 _ ik) I * 2 en Oe ary afi Pf ie He , ; iii - —_ ee { if ! | é «

ry : iV te . + ° his es | i ds : ; f f Fg =~ - : Seti A s* Tr. & ul LO) ppc) oth vi ERD am, hey ny toe © . ‘% P ° 7 o e c Fe OL OUR ee, ab: Gy Deo een

din, Sts te . SP, eT oh 4s. hh SAMOS. a) i «at veg ay ;

4s, fat on i ii) oy [ Ot Geni th Rtee ye :

&

ERRATA.

Page 74, line 27, for m1 ead my

Po Nino 28, Ye ceed a

» line 29, for m8 ead ms:

as

JOURNAL

OF THE

ASIATIC SOCIETY OF BENGAL.

—@—

Part II1.—NATURAL SCIENCE.

No. I.—1887.

T—On the Influence of Indian Forests on the Rainfall—By Henry F. Buanrorp, F. R. S., Meteorological Reporter to the Government of India.

[Received Jan. 20th ;—Read Feb. 2nd, 1887. ] (With a Woodcut.)

The following paper is an extract from the yet unpublished manu- script of the second part of a paper on the Rainfall of India; the first part of which has already been issued as an official publication in the Indian Meteorological Memoirs. In consideration of the great economic importance of the subject treated of, and also with a view to encourage and assist further enquiries, whenever favourable opportunities may offer, it has seemed desirable to publish this discussion in an independent form, and in anticipation of its appearance in the Indian Meteorological Memoirs; where it will form but a small and subordinate part of a memoir dealing with a subject of much wider range.

In an instructive paper originally communicated to Petermann’s Mittesilungen, and subsequently published in translation in the Quarterly Journal of the Royal Meteorological Society, M. Woeikoff appeals em- phatically to the evidence afforded by the Indian rainfall registers, in support of his contention that the action of forests is to increase the rainfall of a country. His appeal is directed chiefly to the contrast afforded by the Assam rainfall with that of the Gangetic valley plain, in about the same latitude and the same distance from the sea; and he

] 5

2 H.F. Blanford—Influence of Indian Forests on the Rainfall. [No. 1,

apparently attributes the great difference displayed by these two pro- vinces, wholly or mainly, to the fact that, while the former is extensively covered with forest, the latter, up to the Terai, is a broad sheet of field cultivation.

In this view I am unable to coincide. Without denying or even questioning the effect of forests as one element of the result, the con- clusion thus formulated seems to me far too sweeping. M. Woeikoff considers, and I think rightly, the action of forests in enhancing the rainfall to be twofold. Firstly, they help to store water, by protecting the soil, and to keep up a constant evaporation ; and secondly, by check- ing and obstructing the movement of the wind, they prevent the evapo- rated vapour being carried away, and tend to produce that calm state of the atmosphere that is favourable to ascending currents and local precipitation. But swamps, such as occupy large tracts of the Assam valley, and the numerous broad river channels that intersect it, must contribute a not unimportant quota to the vapour constituent of the local atmosphere ; and the comparative stagnation of the air in the Assam valleys and the exclusion of those dry westerly winds which play so im- portant a part in the meteorology of the Gangetic plain are certainly due, in far larger measure, to the fencing in of the Assam valley by the Patkoi, Naga, Khasi, and Garo hills, and, as regards Upper Assam, to the interception of westerly currents by the mid-valley obstruction of the Mekhir hills, than to any retardation of wind movement that can be effected by the forests. Furthermore, the action of the surrounding hills in setting up a diurnal convection of the humid-atmosphere, and its consequent dynamic cooling and precipitation, an action which also takes place in the much less humid hill tracts of the peninsula, is a very important item in the causes which contribute to produce the heavy spring rainfall of Assam ; a precipitation not very greatly inferior to that of the summer monsoon. The other or passive effect of hills in enhan- cing rainfall, namely, the forced ascent of horizontal air currents, is less important in Upper Assam (the tract more particularly referred to by M. Woeikoff), although exhibited by the southern face of the Khasi hills, overlooking Sylhet, in a degree without parallel elsewhere in the world. But to the other causes above specified must certainly be attri- buted by far the larger part of that prevailing high humidity and copious rainfall which foster the exuberant vegetation of the province; render- ing it, in the rich variety of its flora and its prolific insect life, compar- able with the teeming productiveness of the Malay region.

The difficulty so conspicuously illustrated in the foregoing example, namely, of disentangling the combined effects of a number of causes all favourable to increased rainfall or the reverse, is one which renders it

1887.] H. F. Blanford—Influence of Indian Forests on the Rainfall. 3

almost hopeless to seek for decisive evidence of the influence of forests by any comparison of the rainfall of different provinces, or of areas suffi- ciently large to display the contrasted effects in a striking and convin- cing manner. The best and perhaps only satisfactory kind of evidence, were it obtainable, would be the comparison of the rainfall of one and the same tract (one of at least some hundreds of square miles in ex- tent) for many years, first while covered with forest, and again for many years after clearing. It is, however, not until a tract of virgin forest has been brought under the destructive operation of civilizing agencies, that, as a general rule, any attempt is made to record its rain- fall ; and when, therefore, the conditions necessary to obtain one term of the comparison are rapidly disappearing. The reversal of this order of things, the conversion of bare or at least partially wasted tracts into protected forest, is one, however, of which India already furnishes some examples, and with the progress of forest protection may yet furnish more ; and if due advantage be taken of these as they present themselves, it may yet be possible to obtain rainfall data which may afford valuable and indeed practically onclusive evidence on the point in question, even if not fulfilling in all respects the rigorous conditions of the logical method of differences.

One instance of the kind, on a scale large enough for all reasonable demand, has lately been brought to my notice by Mr. Ribbentrop, and has been quoted in my Report on the Administration of the Meteorolo- gical Department in 1885-86; and despite some shortcomings in the due verification of the data it furnishes, shortcomings which it is now impossible to make good, it will probably, in the course of some years, as nearly fulfil the conditions of a test case as we are likely to attain to in an experiment of such magnitude. In some respects, indeed, the circumstances of this case are unusually favourable. The vicissitudes of the rainfall of the Central Provinces are smaller, proportionally, than those of any other province of an equally moderate average, and of the 22 stations, the rainfall registers of which will be brought in evidence, not less than 10 are regular meteorological observatories, working under the Meteorological Department of the Government of India.

The region referred to in the Ist part of my Memoir on the Rainfall of India as the Central Provinces south, has been described as a hilly and jungle clad country, including some extensive fertile plains, especially that which surrounds Raipur. The northern portion consists of the range of broken tablelands and hills here spoken of as the Satpuras, and these are largely clothed with forest. According to the most recent report of the Officiating Inspector General of Forests, the area of forest in the Central Provinces is estimated at 54,600 square miles, of which

4 H.F. Blanford—Influence of Indian Forests on the Rainfall. [No. 1,

about nine-tenths are either in or to the south of the Satpura range. The area of the Central Provinces south is about 61,000 square miles, Hence about five-sixths of the whole are under forest. Now, prior to the year 1875, these forests were systematically wasted by the destructive method of cultivation practised by the hill tribes of Gond- wana, as of other wild tracts in India and Burma. It is known under various local names, such as Kumri, or in the Central Provinces déhya cultivation, and is thus described by Dr. Brandis: ‘‘ A few acres of forest are felled one year, the wood is burnt and acrop of grain raised on the clearing; the next year this is abandoned, a fresh piece of forest is felled elsewhere, a crop is raised, and it too is abandoned in its turn; and so on, a fresh clearing being made every year.”’

It will be readily understood how under such a system, in the course of some years, extensive forests may be devastated, even by a _ sparse hill population of nomad habits. And accordingly, in the intro- duction of the Central Provinces Gazetteer, published in 1870, Mr. C. Grant speaks of the state of the forests in the following terms: “‘ The tree forests of the Central Provinces have, however, been so much ex- hausted, mainly owing to the destructive dahya system of cultivation practised by the hill tribes, that, except in one or two localities, the labours of the Forest Officers will, for many years, be limited to guard- ing against further damage, and thus allowing the forests to recover themselves by rest. By far the greater part of the uncultivated lands belonging to Government are stony wastes, incapable of producing a strong straight growth of timber.”

In 1875,* the suppression of déhya cultivation was taken systemati- cally in hand, and in the course of a few years, with such success, that Mr. Ribbentrop writes in 1886, My attention was directed during a recent visit to the Central Provinces, to the extensive growth of young forests in areas formerly under Aumri cultivation. Ten or fifteen years ago, such temporary cultivation was practised throughout the country, and thousands of square miles were thereby laid barren year after year. Since then, this method of cultivation was stopped, and though a great part of the area affected was subject to annual fires, a more or less dense forest growth has sprung up. I concluded that this must have had an influence on the rainfall, sufficiently appreciable to be gauged by meteo-

* I understand from Mr. CG: E. Elliott who worked as Settlement Officer in the Central Provinces prior to 1875, that endeavours were then in progress to check ddhya cultivation, so that the statement in the text which I make on the authority of Mr. Ribbentrop must not be taken as rigorously exclusive. The interpretation of the evidence here adduced will not, however be appreciably affected by this correction,

1887.] H. F. Blanford—Influence of Indian Forests on the Rainfall. 5

rological records. The results gathered from such records are beyond expectation and shew that, with the exception of stations situated in the cultivated valley of the Nerbudda, a steady increase of rainfall has taken place during the last ten years, and, as might be expected, especially during the last period of five years.” .

In dealing with the evidence of the rainfall registers, I shall, in the first place, compare the averages-of the 9 or in some cases 10 or 11 years ending with 1875 (the year in which the suppression of the dahya cultivation is stated to have been taken in hand) with those of the ten subsequent years, 1876—1885; and this I shall do separately for the stations within the area immediately affected by the forest preservation and for those at a greater or less distance therefrom. These latter are Saugor and Damoh, the forests near which have not been frequented by dahya cultivators, or which are surrounded by native states in which no change of system has been attempted, Jubbulpore, Narsinghpur, Ho- shangabad, and Khandwa in the fertile and highly cultivated valley of the Nerbudda, and where the tendency of late years has been towards an extension of permanent cultivation, and Raipur in the centre of the great wheat-growing district of Chattisgarh.

Comparison of the Average Rainfall of 9 to 11 years of Déhya Cultiva- tion with that of 10 years of Protected Forests.

A, In affected areas. Forests ; Forests I unprotected. | protected. | “Crease. STATIONS, ! Inches: Inches Period. | Rain- ; Period. |} Rain- | Inches. fall. } fall. Badnur 5 os ... | 1867-75 | 39°83 | 1876-85 | 47°83 | + 8:00 Chhindwara : YS ... | 1865-75 | 41°43 | 1876-85 | 48°48 | + 7:05 Seoni mae ... | 1865-75 | 52°07 | 1876-85 | 54°76 | + 2°69 Mandla ; sah ... | 1867-75 | 53°58 | 1876-85 | 56°32 | + 2°74 Burha ; ie ... | 1867-75 | 64°51 | 1876-85 | 71°65 | + 7:14 Bilaspur : vee ... | 1865-75 | 41°85 | 1876-85 | 54°81 | +1296 Sambalpur wa ase ... | 1867-75 | 54°80 | 1876-85 | 67°93 | +1313 Dhamtari ; a ... | 1867-75 | 48°83 ' 1876-85 | 46°90 | 1:93 Bhandara a ... | 1867-75) 49°90 | 1876-85 | 57°79 | + 7°89 Nagpur re, : ... | 1866-75 | 41°54 | 1876-85 | 51°85 | +10°31 Wardha os - ... | 1866-75 | 36°10 | 1876-85 ; 46°63 | +10°53 Brahmapuri es ... | 1867-75 | 53°95 | 1876-85 | 57°48 | + 3°53 Chanda “f “a ... | 1866-75 | 47:14 | 1876-85 | 54°29 | + 7:15 Sironcha os ; vw ) 1867-75 | 44°17 gene 48°38 | + 4°21 | Mean | +

6 H. F. Blanford—Injluence of Indian Forests on the Rainfall. [No. 1,

B. In unaffected areas.

Saugor Sa Soc ... | 1866-75 | 55°97 | 1876-85 | 40°62 { —15-35 Damoh vee eee ... | 1867-75 | 54°76 | 1876-85 | 46°82 | 7:94 Jubbulpore eee vee ... | 1866-75 | 60°66 | 1876-85 | 56:28 | 4:38 Narsinghpur a ... | 1866-75 | 55°46 | 1876-85 | 50:40 | 5-06 Hoshangabad__... oe .-. | 1866-75 | 47-08 | 1876-85 | 57°73 | +10°65 Khandwa eee one .. | 1867-75 | 34°74 ) 1876-85 | 33°32 |! 1:42 Raipur eee ees .. | 1866-75 | 51°59 | 1876-85 | 54°47 | + 2-92

| Mean | 2°94

The contrast, thus shewn, is sufficiently striking; but taken as they stand, it can hardly be said that the figures do more than afford a certain presumption in favour of the view that the difference shewn by the two series of stations is to be attributed to the preservation of the forests. In the first place, as I shall shew elsewhere, the probable error of a ten years’ rainfall average of a station in the Central Provinces is about 5 per cent., and this may be either in excess or defect. In the extreme case of the errors being in opposite directions in the two decen- nial periods compared, the greater part of the apparent increase of list A would vanish. And, in the second place, the majority of the stations in the second list lie to the north of the Satpura range, those of the first list either on the range itself or to the south of it; and, as this range about coincides with the southern margin of the tract commonly followed by the cyclonic storms of the summer monsoon, the distribution of the rainfall might be much affected by the fact of a series of such storms following a more southerly or more northerly path, or by the western branch of the monsoon, which brings nearly the whole rainfall to the region south of the Nerbudda valley, being in several years, relative- ly to its normal average, stronger and more rainy than the eastern branch, which contributes to the rainfall north of that river.

But there is another way of dealing with the facts which will not be open to such objection. Any effect really due to forest protection must necessarily have been progressive. Some few years were passed in inducing the jungle tribes to take to settled cultivation; again, the reproduction of the forest growth on the tracts formerly denuded is a process requiring many years for its accomplishment; and, finally, the protection of the forests from destruction by annual fires in the dry season has been steadily extended year by year. If, then, it should appear, on comparing the rainfall of the affected tract in successive years, that the increase has been steadily progressive and on the whole in a degree commensurate with the average difference of the two decen- nial periods above compared, the probability of such increase having been brought about by the protection of the forests will be enormously enhanced.

1887.] H. F. Blanford—Influence of Indian Forests on the Rainfall. 7

The data for this comparison are afforded by the following table which exhibits, in the second column, the mean rainfall of the 14 stations, enumerated in the A list of the previous table, in each year from 1867 to 1885. The third column gives what may be termed progres- Sive averages. Hach average is that of 5 years, obtained by the formula,

at46b+6c+4d+e , A et ear eae

wherein a, b, c, d, e, represent the mean rainfall in any five conce- cutive years and c, the progressive average for the third year of the series. As a standard of comparison, I give, in the fourth column, the average rainfall of the whole Indian area (with the omission of unrepresented tracts). The average rainfall is taken at 42 inches. Lastly, the fifth column shews the progressive averages of the rainfall of India computed from column four.

Comparative Table of the Mean Annual Rainfall of the Forest Region in the Central Provinces and of India from 1867 to 1885.

Central Provinces. India. Year. Condition. Annual |Progressive} Annual Progressive mean average. mean. average. Inches. Inches. Inches. Inches. 1867 | 55:08 sipptilbe 44:8 40°2 68 if ENGL us Sealer Se 35°4 40°3 69 B 47°97 45°28 42:4, 41-0 "0 ‘a 50°42 47°71 43°5 42°6 val es 45°52 48°45 42°9 43°0 "3 & 53°31 47°47 44°3 41:7 r3 qi 39°18 47-02 37°5 42:2 4, sa 50°48 48°85 46'6 424 "5 56°60 50°15 44:4, 42°4, "6 42°32 49°58 37°5 40'5 "7 % 52°50 50°40 37°7 41:1 "8 52°47 52°60 48°3 43°3 "9 55°67 53°85 43°7 43°5 80 3 51°83 54:50 40°4 424, 81 = 57°90 65°31 42:1 424, 82 a. 54°22 56°52 44:6 43-0 83 iS) 57°73 58°57 41‘9 43°1 84, Pa 64°63 Sh a Ein ase pig 85 | 57°43 eee eee 43°1 eee ree

8 H. F. Blanford—Influence of Indian Forests on the Rainfall. [No. 1,

The variations exhibited in this table are represented graphically in the following figure.

Gueeccseegele coisa Eta Buen al ae ES : Pek ed

AN EprLL Ae

| A AG A (5a Se SASSER

+ le, [|

RV

aes ee ao do Progr. av.

Now the third column of this table shews, not only that the in- creased rainfall of the protected forest region has been on the whole progressive since 1876 (the year after protection was systematically enforced), but that its progression has been commensurate with the increase of the decennial average shewn in the previous table; a very important point. As compared with the general average of the pe- riod antecedent to 1875, a rainfall of 48 inches, in integral figures, had risen to 58 inches in 1883, an increase of more than 20 per cent. Whether this increase will be sustained at its full amount by the results of future years is, however, very questionable. The rainfall of 1884 was extraordinarily high, and whereas, as may be observed in the graphic representation of these changes, the rainfall of the Central Provinces rises and falls pari passu with that of the whole of Iudia in a somewhat remarkable degree (having regard to the comparative smallness of its area), the pregressive average rainfall of India as a whole for 1883 was nearly 3 per cent. above the general average between 1867 and 1875. But after making all due allowances, in so far as any legitimate con- clusion can be drawn from the experience of the last ten years, it would seem that, owing to some local cause, the mean rainfall of the afforested region of the Central Provinces here considered, an area of nearly

1887.] H. F. Blanford—Influence of Indian Forests on the Rainfall. 9

50,000 square miles, has been increased in a very remarkable degree, and I am unable to assign any other probable cause for this than that of the protection and consequent restoration of the formerly wasted forests.

The evidence thus afforded in favour of the influence of forests on rainfall appears to me to be of considerable weight and importance, in virtue both of the magnitude of the area yielding it and of the apparent distinctness of the result. With one exception and one only, it fulfils all the conditions of a rigorous test case. The area is one and the same ; the history of the changes to which it has been subject are definitely and accurately known; and, as will be shewn elsewhere, the rainfall registers, if but few in proportion to the area, are sufficient to afford a datum the probable error of which is small in comparison with the mag- nitude of the effect shewn. The only remaining points to which excep- tion may conceivably be taken are the trustworthiness of the records used, and the sufficiency of the periods compared to yield valid averages.

On the first of these points, I can add but little to what has been already written in the introduction of Vol. III, Part I, of the Indian Meteorological Memoirs. Speaking from recollection (for I have been unable to obtain the desired verification of the fact from official records), I believe that new rain-gauges, of Glaisher’s pattern, from one of the principal London makers, were furnished to all the stations the registers of which are here dealt with, about the year 1867, at all events before 1870, that is to say, at or near the beginning of the period for which the registers are complete, and there are therefore no grounds for suspecting that the increase of the registered rainfall during the last ten years has been influenced by a change in the instruments used. And this is the most important consideration. With respect to the registering agency, as far as I have information, it has been the same throughout. Dr. S.C. Townshend, who was Sanitary Commissioner of the Central Provinces, and who in 1868 established the observatories, which, in 1875, were incorporated in the Imperial system, took much personal interest in all the meteorological work of the province, and there is no doubt that his action was attended with beneficial results. But this change, like that of the instruments, dates from the beginning of the period now under consideration ; at all events from 7 or 8 years anterior to 1875.

On the second point, namely, the sufficiency of the periods compared to yield valid averages, I have ascertained that a ten years’ register of the Central Provinces’ stations, Jubbulpore and Nagpur, has a probable error of 5 per cent., namely, in the case of Jubbulpore of 2°7 inches, in that of Nagpur of 2:2 inches, and these may be taken as fairly illustrative examples of the whole province. These, however, are the probable errors of individual stations, and, as will be further demonstrated elsewhere, the

2

10 H. F. Blanford—Influence of Indian Forests on the Rainfall. [No. 1,

mean rainfall of a whole province is much less variable than that of a single station. For, if we take the average of either the first ten years or the last ten years of the figures in the third column of the table, on page 7, we have an average of 1867—1876, 47°45 ins. probable error + 1:56; average of 1876—1885, 54°67 ins. probable error + 1°22, which is but little more than half the probable error of either Jubbul- pore or Nagpur for an equal period. This is small in comparison with the difference of the two averages, namely, 7°22 inches. Assuming the extreme case, that the first average is 1°56 inches below the real mean and the second 1°22 inches about it, these differences being both due to fortuitous and not steadily progressive causes, there would still remain 444, inches of increase unaccounted for. This is perhaps not such as to warrant conviction that the average rainfall of the Central Provinces south of the Nerbudda has really increased by that amount; still less ~ does it warrant the positive assertion that such increase, assumed as real, is due to the preservation of the forests; but at least, in so far as any inference is admissible from such data, the evidence seems to afford much support to that view.

Direct observations of a character similar to those of Prof. Ebermeyer in Bavaria, namely, comparative measurements of the rainfall at pairs of stations near the margin of forests, the one within, the other without the forest, have been carried on in Dehra Din and Ajmere, during the last year or two, by officers of the Forest Department. Some of the results of these were given in the Administration Report of the Meteo- rological Department for 1885—86, and I have since visited the Dehra Dun stations and some of those in Ajmere. In the case of the former, the conditions are satisfactory, in so far that the forest on the site of the observatories is a vigorous growth of chiefly Sal coppice* with a well- defined boundary, and the observatory stations are, in the one case, well within the forest, in an opening only just large enough to prevent the gauge being sheltered, or its contents unduly added to by the drip of the trees; in the other, in an open maidan of coarse grass and scrub, with only a rare tree here and there. But the interval between the two

* As testifying to the importance of this condition, I extract the following from a letter lately received from Dr. D. Brandis, for many years Inspector-General of Forests in India; “I would draw your attention to a point which I used to urge in India, whenever I wrote on the subject; viz., that forests, in order to exercise an effect (on the rainfall), must be dense, and must not consist of a few bushes and trees, here and there. Fire protection alone has the effect of making the forest grow up dense, and I am disposed to think that the large extent of fire

protected forest in the Central Provinces may, in course of time, affect the rain- fall.’

1887.] H. F. Blanford—Influence of Indian Forests on the Rainfall. 11

stations of each pair is hardly enough to shew the full influence of forest in the one case, or to exclude it in the other; and it can only be expected that, under such circumstances, any difference depending on that influence will be very small.

There are two such pairs within about 6 miles of Dehra Din, on the skirts of the Sivalik forests, the one at the Ramgarh, the other at the Rajah’s forest. In the case of the Ramgarh forest, at which the obser- vatories have been longest in existence, the two observatories are 750 yards apart; the outer 400 yards from the forest boundary, the inner 350 yards from it. At each station, there are two rain-gauges, the one on the ground, the other at a height of 60 feet, being perched on the summit of a scaffold, which raises it above the tops of the neighbouring trees. The rain-gauges are Symon’s pattern, 5 inches in diameter, and the measurements are all made with the same measure glass. The observer has been regularly trained in his duties (which include keeping 4 registers of temperature and humidity, under corresponding condi- tions), and his work seems to have been regularly performed. The results for the years 1884 and 1885 are given in the following table :—

Lower gauges. Upper gauges.

Months. : . Outer. | Inner. ra Outer. | Inner. a

ee |

June 1884 ... 5 3°66 4:07 3°88 | +027 July Sf) aclu! Sauk] MEGS )) 26746 26°44, | +1°72 August ore aa 21°18 21°74 21°23 +1°35 September , .. ‘Pe 17°53 18°78 18°01 | +082 October Risky Gun BEE 0°28 0°39 0°37 | +011 November _,, ais sas 0 0 0 0 December ,,_... = 0 0 0 0 Total 68°29 71°44 69°63 | +4°27 January 1885 ... Py) 4°20 448 | +0°28 4°56 4°63 | +0°07 February ,, ane én 0°85 0-70 | —O15 0-77 0°67 | —0:10 March - tae ans 0°48 0°39 | —0°09 0°42 0°36 | 0°06 April Bae iy. pain 1 Seah 9 OBER oh OeRt 0-45 | 050] 40:05 May ea cys aed. abe Ee |) ag GOA tO Get 606] 6791 +073 June eon EOS! toe | Ob 975 | 1061 | +0:86 July y aae sos 9°81 9°90 + 0°09 9°27 9°88 + O°6L August a ee soe 44°64 44°91 + 0°27 43°56 44°45 +0°89 September ,, _... aan 6°24 551 | 0°73 6:06 547 | —0°59 October ‘i ake ave 0) e) 8) 0 8) a) November ,, ae Jen 0 0 0 0 0 0 December ,, Ain eat 3°45 3°49 | +0°04 3°48 3°52 | + 0°04: Total ...| 85°77] 8688} +091 ] 8338] 85:88 | + 2°50

ee

12 H.F-. Blanford—Iufluence of Indian Forests on the Rainfall. [No. 1,

The observations at the Rajah’s forest extend over a shorter period. The stations are less than a mile distant from the former, and the arrangements are similar; the surrounding conditions of each of the pair being being strongly contrasted. The outer observatory is 1,750 yards from the forest boundary, the inner 1,000 within the forest, which is of the same character as the Ramgarh forest.

Lower Ganges. Upper Gauges. Months. . Outer. | Inner. aie Outer. | Inner. Spe

March PESO sss waa P 0°27 P 0°21 0°23 | +0°02 April ay lei LA 0°06 0°42 | +0°36 0°32 0°36 | +0°04 May nel hos a 4°69 3:99 | —0°70 4°36 4:04 | 0°32 June daphne ne 10°47 11:70 | +1:23 10°07 11°42 | +1°35 July eattee™, we ve 9°81 10°63 | +0°82 9°47 9°88 | +0°41 August ae Whine sie 47°50 45°87 | 1°63 46°99 45°87 | —112 September ,, we nae 2°43 2°46 | +0°03 2°40 2°41 | +0°01 October pes S Yh 0 0 0 0 0 0 November ,, “ae 7 0 0 0 0 0 0) December ,, «. ca 3°40 3°54 | +0°14 3°43 3°45 | +0°02

Total ie 78°36 78°88 | +0°25 77°25 77°66 | +0°41

Tn this case, while, in most months, the rainfall at the inner station is appreciably higher than at the outer station, as shewn both by the elevated and ground level gauges, this gross excess appears to have been nearly neutralized by falls in May and August, which were in excess at the outer station. The result of the evidence is therefore doubtful. But in the case of the Ramgarh station there does appear to be a decided balance of rainfall in favour of the inner station.

I do not give the results of the Ajmere observations, because the difference of the conditions within and without the boundary of the forest, as far as I have seen them, depend much more on the form and slope of the ground than on the density of the forest growth, and I do not think the comparative observations have much bearing on the question at issue. :

There remains one case which, although dependant on purely arti- ficial conditions, might yet afford evidence of some weight in connection with the present subject, could we only be sure that the observations had been taken with the care and precaution indispensible to any valid comparison.* In the very heart of the plain between the Ravi

* For the following information I am indebted to Col. Home, R.E., late ‘Secretary to the Punjab Government in the Irrigation Department of the Public

1887.] H. F. Blanford—Influence of Indian Forests on the Rainfall. 13

and the Jhelum (two of the five rivers of the Punjab), and about 50 miles to the south of Lahore, a vigorous forest has been established, by planting and irrigating the planted land from the Bari doab canal. The forest area covers 315 square miles and has now been established 16 years.* Outside the forest and to the east and south-east are lands which are cultivated, also with irrigation from the canal; and on the margin of this tract, four miles from the forest, is the small civil station of Chunian. Since 1864, a rainfall register has been kept regularly at Vahn (within the forest, half a mile distant from the nearest forest boundary), and also at Chunian; and since 1870, a third register has been kept at Bhambeh, a station on the Bari doab canal, in a position very similar to Chunian, but 13 miles to the north-east of the forest boundary and 19 miles north-east from Ghanga Manga or Vahn.

The rainfall chart of the Punjab shews that, in this part of the province, there is a steady increase of rainfall in a north-east direction or from Chunian to Bhambeh; steady, that is to say, apart from the influence of purely local conditions, and therefore, were the whole surface of the tract such as it is immediately around Chunian and Bhambeh, it might be anticipated that the mean rainfall of any inter- mediate station should be intermediate between those of Chunian and Bhambeh, in inverse proportion to their respective distances. The mean rainfall of Bhambeh, deduced from 17 years’ registers, is 17°27 inches; that of Chunian, deduced from the same period, is 14°05 ins. If, then, Vahn, which is 19 miles from the former and 63 miles from the latter station, had a rainfall intermediate between the

Works and now Secretary to the Government of India. ‘“ Two gauges are placed side by side, the receivers are 43 feet above the ground. One is an ordinary tube gauge, measurements made with a graduated rod. The other a Watsen’s continuous self-registering gauge, which is taken to pieces, cleaned and re-adjusted on the 1st April yearly. The bearings of the gauge are silver plated copper tubes and with very ordinary care in adjustment, they register very correctly. Instruc- tions about registering rainfall are very distinct and I believe they are obeyed.”

| * Mr. H. C. Hill, Conservator of Forests in the Punjab, writes, Changa Manga is a compact block of 20,242 acres, of which 8,399 are wooded with planted Sissoo (Dalbugia Sissoo). The remainder is under ordinary scrub. The age of the plantation dates back to 1866-67, but little was done for 3 years and the age of the forest may be taken as 16 years. The trees (excepting those in the canal avenue averaging 63 feet) of our best compartments average 50, 51 and 53 feet in height and all compartments have an average of 40 or more.”

“The watering of the forest begins in April and goes on more or legs till September. Very little of it ever gets a second watering in the year, but that given is a good soaker of 3 or 4 feet depth of water. The ground to the east and _ south except where 2 rahhs are touched, is all under cultivation and irrigated. Irrigation mostly from June to April,

14 H. F. Blanford—Inflwence of Indian Forests on the Rainfall. [No. 1,

above amounts, in inverse proportion to the distances of the ‘two stations, the average of the same 17 years would be 14°85 inches. It is actually 15°76 ins., or nearly 1 inch above the computed pro- portion.

I am far from considering this result as conclusive on the point at issue. In some years,