554 



NA TURE 



[February 13, 1908 



Allahabad is situated about 300 feet above sea-level at 

 the junction of the Ganges and Jumna, where their general 

 direction is changing from E.S.E. to E. The plain of 

 these rivers forms part of a great plain 1300 miles long 

 and about 200 miles broad, in no part of which does the 

 height above sea-level exceed 1000 feet. On the north it 

 is bounded by the Himalayas, which change their direc- 

 tion from N.W. to W. in passing from the Punjab 

 to Assam. Near Allahabad the direction is appro.ximately 

 W.N.W. 



To the south the ground rises gradually to the plateau 

 of Central India, across which runs a low range of hills 

 from Bombay in an E.N.E. direction, passing about 150 

 miles south of Allahabad. 



Allahabad is therefore near to the south edge of the 

 flat bottom of a trough with sides converging towards the 

 east, the south side being very slightly inclined and of 

 small elevation compared with the north. Lucknow is 

 no miles N'.W. of Allahabad, and lies at the centre of the 

 flat bottom. The motion of the air in such a trough is 

 complicated, but the general result is that air flowing in 

 or out transversely is deflected towards the right in its 

 course, the effect being in both cases to produce motion 

 parallel to the trough. On these motions will be super- 

 posed the effect of the general circulation of the atmo- 

 sphere, which is both actually and theoretically westerly 

 in the upper regions so long as the trough is definitely 

 north of the thermal equator, becoming easterly when the 

 thermal equatorial region includes the trough. 



The exposure of the anemometer at Allahabad was 

 excellent, but at Lucknow was not so good, and at the 

 latter place the instrument, during the later years, was 

 not kept in proper working order. It is probably partly 

 due to these causes that the records from Lucknow show 

 winds considerablv weaker than those from Allahabad, 



The results of the records have been arranged in tables 

 giving for each month for each hourly interval of the 

 day (I) the mean movement of the air, irrespective of 

 direction: (2) the number of winds recorded under each 

 octant of the compass ; (3) the number of miles recorded 

 under each octant of the compass ; (4) the mean coordinates 

 of the resultant wind movement. 



For exhibiting the leading features of the air movement 

 these results have been charted, and a series of carefully 

 drawn plates is given at the end of the discussion. In 

 addition 10 w'ind roses, showing the amount of wind in 

 each direction and the proportion of calms, there is an 

 excellent set of diagrams showing for each month of the 

 year the diurnal variation of the air movement and the 

 mean monthly resultant velocity. The diurnal variation 

 bears no direct relation to the ordinary diurnal pressure 

 variation. 



For the purposes of discussion, Sir John Eliot divides 

 the year into two periods, the dry season extending from 

 the middle of October to the middle of June, and the wet 

 season during the remaining four months. The dry season 

 is further subdivided into the dry cool season, November to 

 February, and the dry hot season, March, April, May. 



During the cool season, pressure gradients are small, 

 and the main feature of the distribution is the persistent 

 continental high pressure. Locally, the isobar through 

 Allahabad at 8 a.m. in January runs nearly through 

 Lucknow in a N.W. direction, becoming more northerly 

 as the day advances, while the wind changes its direc- 

 tion from W.N.W. to N.N.W. Thus there appears to be 

 a correspondence between the wind direction and the local 

 pressure gradients similar to that noted every day on the 

 synchronous weather charts for temperate latitudes. The 

 variations are, however, too rapid for the development of 

 the full effect of the earth's rotation in producing an 

 approach to parallelism between the isobars and the wind 

 direction. 



It is important in considering this rotation effect to 

 remember that it acts as a modifying influence in con- 

 junction with the pressure distribution, and although the 

 latter is in the end the outcome of the air motion and 

 temperature variation, there is nothing to warrant the 

 assumption that the combined effect on air motion is to 

 produce always a veering in the wind. Air starting from 

 rest and moving across a permanent system of isobars 

 will veer as it progresses, but a change in the direction 



NO. iggS, VOL. 77I 



of the gradient may more than counteract this action of the 

 earth's rotation. 



Briefly, if the pressure fall in unit distance along two 

 perpendicular straight lines, O.v, Oy, by amounts a, ;3, and 

 if the resultant velocity due to the effects of the pressure 

 gradient, friction, and the earth's rotation be proportional 

 to the gradient and make a constant angle ft with the 

 isobars, the components of velocity in the two directions 

 Ox, Oy, wm be />-(acos e-;3sin fl), -fe(a sin 9-1-/3 cos 6). 



If, now, a, /3 are the mean values of o, deduced from 

 inean pressure distribution, and if h, v are the mean values 

 of the components of wind velocity, we find 



ti = fe{acose-;3 sin «), ~c= -fc(asin fi-l-;8cose) 

 and the same relation, therefore, holds for mean values 

 as for synchronous distributions. The angle * depends on 

 friction and on the time the motion has been in progress ; 

 u and V will therefore vary between limits depending on 

 these factors. The general relation is, however, simple, 

 and it appears desirable to test its applicability to motion 

 in the large unbroken plain, ample and suitable data for 

 which are furnished by the present series of memoirs. 



An examination of the diagrams shows that at both 

 .\llahabad and Lucknow the cold season has the greatest 

 percentage of calms and the smallest air movements. The 

 latter are, however, steadier than at any other season. 

 Calms are ih times more frequent at Lucknow than at 

 Allahabad, the average number at the former place being 

 30 per cent, of the total number of observations. The 

 mean direction of the air movement is slightly \V. of 

 N.W. at both places. The diurnal variations of magnitude 

 are similar at both places, the maximum being reached 

 about 3 p.m., when the average velocity is more than 

 double that of the evening. The changes in direction at 

 the two places are very different. The wind usually veers 

 throughout the day at .Allahabad, and backs during the 

 night ; at Lucknow the main feature is considerable, back- 

 ing from II a.m. to 3 p.m., and slow veering for the 

 remainder of the day, with slight and very irregular 

 movements at night, the changes being much less than 

 those at Allahabad. 



In the dry season the winds are of maximum intensity 

 at Lucknow and of mean intensity at Allahabad. At both 

 places the actual resultant air inovement is a maximum 

 for the year. The winds are relatively very steady in 

 March and April, and very unsteady in May. The mean 

 direction changes from N.W. to N. at Allahabad, and 

 from W.N.W. to N.W. at Lucknow during the course 

 of the season. The diurnal changes are similar to those 

 of the cold season, but are more marked, and in May the 

 changes in direction arc greater at Lucknow than at 

 Allahabad, but still take place in the reverse direction. 



The winds during the wet season are remarkable for 

 their increased variability in direction. The actual ampli- 

 tude of the diurnal variation of magnitude is considerably 

 less than for the dry season, the winds being less feeble 

 during the night and of average intensity in the day. At 

 .Mlahabad the mean direction of air movement in July is 

 from W.S.W., but during the early morning hours it is 

 nearly S., and at 4 p.m. it is N.W'. by N. At Lucknow 

 the mean direction is N.E., and the variations are less 

 marked, but there is a very remarkable change from E. 

 bv N. to N.E. by N. between 10 a.m. and 11 a.m. The 

 motion is the same as if the places were in a trough of 

 ascending motion the axis of which moved towards 

 .Allahabad in the course of the day. 



Diagrams are also drawn to show the variations of air 

 movement along, and perpendicular to, the axis of the 

 trough, appropriately called the axial and transverse 

 variations. The axial variation shows similar features 

 throughout the dry season. There is a fairly rapid in- 

 crease in the daytime until 4 p.m., after which there is 

 a rapid decrease. During the night there is practically no 

 change. In the wet season the increase is much smaller 

 at both places. At Allahabad the maximum is reached at 

 II a.m., and the decrease takes place slowly during the 

 remainder of the day. There is a feeble secondary maxi- 

 mum at 2 a.m. At Lucknow the increase takes place 

 slowly and irregularly from midnight to midday, while 

 there is a similar decrease until 10 p.m. The transverse 

 variation is throughout markediv different at the two 



1 



