428 



NA TURE 



\_August 30, 1883 



of examining the map now in Stockholm, he is convinced that it 

 is simply a Dutch sea chart from the beginning of the seven- 

 teenth century, and of no value whatever, which he b-lieves 

 Baron Nordenskjold did not, under a mere cursory examination, 

 discover. In consequence of the opinion pronounced hy our 

 correspondent, the Swedish Geographical Society has decided 

 not to have facsimiles of the map taken. 



INDIAN METEOROLOGY^- 

 II. 



"THE title of Paper IV.— "Storms in Bengal, accompanied 

 -* by increased Atmospheric Pressure and the Apparent 

 Reversal of the Normal Diurnal Oscillation of the Baro- 

 meter," by Prof. J. Eliot, — must necessarily appear some- 

 what strange to those accustomed in our latitudes to the frequent 

 masking, if not actual reversal, of the normal diurnal O'cillation, 

 by the large and rapid non-perioJic oscillations to which the 

 atmospheric pressure is subject. 



There are two reasons why this reversal should be rare in the 

 tropics, and of such frequent occurrence, as to be mire often the 

 rule than the exception, in higher latitudes. One is, that the 

 range of the diurnal barometric o-xillation is greatest near the 

 equator, and diminishes as we approach the poles, and the other, 

 that the range of the non-periodic oscillations varies in precisely 

 the opposite way, increasing very nearly in the ratio (as Ferrel 

 has shown) of the square of the sine of the latitude to unity. 2 



Instances of such inversion, are said by Prof. Eliot to be 

 extremely rare in Bengal, but a perusal of the paper leads us to 

 conclude that it is raiher a ca=e of de non apparcntibus than 

 de non existentibus, and that a tendency towards reversal takes 

 place to some extent in all north-westers and analogous storms 

 of a sudden and violent character. 



Humboldt, Col. Sytes, and Allan Broun, have all graphically 

 described the regular march of the diurnal barometric oscillation, 

 but their observations were mostly made a good deal nearer the 

 equator than Bengal, and thus in regions where reversal of the 

 diurnal oscillation would be a still rarer phenomenon. 



In the cases ci'ed by Eliot, including one observe! by Hill at 

 Allahabad, the following characteristic changes were observed 

 about the time of reversal : — 



1. A marked rise of the barometer. 



2. An equally marked and simultaneous fall of the thermo- 

 meter. 



3. A sudden decrease of the tension of aque us vapour. 



4. An instantaneous change of wind direction. 



A consideration of all these features, leads the author to con- 

 clude that in these cases a downru-h of cold air belonging to an 

 upper current (which is known to travel seawards in Bengal, and 

 is therefore dry) takes place in the centre of the area of low 

 pressure belonging to the storm. On no other hypothesis, does 

 it seem possible to explain all the facts, especially the rise in the 

 barometer, and the fall in vapojr tension. 



A close study of these north-we ters, whether acco r.panied by 

 reversals or not, and their analogues in other parts of the world, 

 is certain to unravel much of the complexity surrounding such 

 and all aerial disturbances, and as all facts bearing on them are 

 valuable, and the present writer was for some time resident in a 

 locality (Dacca) where they occur with marked intensity, he may 

 perhaps be allowed to remark that one of the most peculiar 

 features he has noticed in connection with them, is the almost 

 instantaneous return of the wind to its original direction after the 

 rear of the storm-cloud has passed the zenith. Both before and 

 after the storm, the wind in Bengal blows from the direction of the 

 sea (south-east). The storm-cloud appears to form in the north- 

 west by an aggregation of vapour that is carried thither in cloud- 

 lets from the sea. After a time, the threatening mass advances 

 towards the sea, the sea-wind meanwhile blowing towards the 

 advancing cloud with increased force, until the latter has arrived 

 pretty close, when a lull takes place, after which the wind in- 

 stantaneously changes to the opposite quarter (generally through 

 the north), from which it blows with great violence. Then come 

 the characteristics already noticed, together with continuous 

 lightning and hail, the latter often very large. 



When the storm-cloud, the rear of which presents a very 



1 Concluded from p. 407. 



^ "Relation between the Barometric Gradient and the Velocity of the 

 Wind," by W. Ferrel. American Journal 0/ Science, vol. viii., November. 

 1874- 



definite outline, passes the zenith, the wind invariably returns to 

 its former direction, and gradually dies down as night approaches. 

 The whole phenomenon appears to favour the notion which has 

 always been entertained by tl e writer, and is merely a slight 

 extension of the explanation given by Prof. Eliot, that a sudden 

 oblique uprise of moist hot air takes place, deflecting the upper 

 current into a corresponding oblique downward course, which 

 determines the direction of the storm and continues as long as 

 the uprising air interferes with its regular motion parallel to the 

 surface. This explains why the storm always follows the course 

 of the upper current, as well as the immediate readjustment of 

 the original conditions, as soon as the region of ascending air 

 which causes the deflection has passed the spot. 



Paper V. "On the Rainfall of Benares in Relation to the 

 Prevailing Winds," by S. A. Hill. — The observations utilised in 

 this paper, as far as the velocity of the wind is concerned, raise 

 a question of general importance, and one which we think ought 

 to engage the attention of all thoughtful, and certainly all 

 practical meteorologists. Up to 1872, the anemometer was only 

 15 feet above the ground, but in that year it was raised to a 

 height of nearly 80 feet. How what are we to think of the 

 effect of such a change of position on the observations ? 



According to Mr. Stevenson, 15 feet is the lowest elevation at 

 which an anemometer should be placed, since below this height 

 the velocity is found to be enormously affected by the nature of 

 the surface. On the other hand, 80 feet is an elevation which 

 would not only cause the instrument to register a considerably 

 higher velocity, 1 but also secure for it nearly complete immunity 

 from the disturbing influences which would be sure to affect it in 

 the lower position. 



It is indeed very much to be regretted that in setting up ane- 

 mometers in India no sort of uniformity seems to bax'e been 

 attempted. Thus from a list of their elevations above the ground 

 given in the "Meteorological Report for India " in 1S76, every 

 variety of height imaginable occur-, ranging from 5 leet 7 inches 

 at Khandwa to 76 feet 11 inches at Benares ! At no two sta- 

 tions are the anemometers at the same level, and though it is 

 somewhat complacently admitted by Mr. Blanford "(hat it can 

 hardly be affirmed that in the majority of cases the anemometric 

 records are strictly comparable," one is naturally inclined to ask 

 why the instruments could not have been placed, if not exactly, 

 at least more nearly, at the same level. Such an arrangement 

 would seem to be a cardinal requisite where such a sensitive 

 element as air motion is involved, and indeed Mr. Stevenson's 

 experiments have shown that » hile observations below 15 feet 

 are almost useless, the velocity increases rapidly with the height 

 through the first 100 feet ; so that until a correction is made for 

 the height, it would be vain to attempt a comparison of observa- 

 tions made at lower, with those at higher levels. 



Fortunately in this case the discussion mainly relates to the 

 direction of the wind, so that the conclusions are not seriously 

 affected by the change in the position of the instrument. 



The chief conclusion arrived at by the author is similar to that 

 obtained by Mr. Blanford in the first paper of this series for 

 Calcutta, except that Benares affords no corroboration of the 

 conclusion drawn from the former register, that "rain is the 

 more probable in proportion as the deflection of the monsoon 

 current is greater." 



It appears, nevertheless, that there is a well-marked connec- 

 tion between the amount of rain that falls in a day and the 

 deflection of the rain-bearing current, the maximum amount 

 being from the opposite quarter (north-west) to that from which 

 the monsoon blows. 



Paper VII. "Variations of Rainfall in Northern India," by 

 S. A. Hill. — This is one of the most interesting and important 

 papers of the whole series, its ostensible object being partly to 

 test the soundness of the idea which was propounded by Dr. W. 

 YV. Hunter and others in 1S77, that sunspots, rainfall, and the 

 occurrence of droughts and famines were closely associated in 

 India. Regarding this vexttta qucestio it may be said at the out- 

 set that while the general results of an investigation embracing 

 an area which covers eleven degrees of latitude and twenty of 

 longitude (equal lo that of the British Isles, France. Germany, 

 Austria, Holland, and Belgium combined), like those of Mr. 

 Blanford for Southern India, bear out Meldrum's theory of an 

 eleven-year cycle of raiufall, they exhibit certain irregularities, 

 or, more properly speaking, double oscillations, which, as Prof. 



1 According to Mr. Stevenson's formula, which holds near the surface, the 

 velocity would be increased by this change of position in the ratio 1*33 : 1. 

 Thus for 1S72 the observed mean value was 67 8. To make this comparable 

 v. Lib the years that follow, it should be So^ ! 



