''! 



TIIANSACTIONS OF SECTION A. 



039 



.'liiin' 



can he 

 jiivius to 

 I surfaco 



of time 



(half an hour) roquirud for poliHliini,' wlicn litlli) alteration of fi(,'uro Id required, 

 and when no scratches Imvo to bo worketl ont. 



lo. .1)1 Accnitnti'f some prtjliniuiftrj/ E-rjicrinieiiL-i with Jliiuiiii'i^ Atip.Diomders 

 iitlechrd to Kite-wir<is.^ Hi/ Prolossor K. Douglas Akciiiuaij), 2r.A. 



In this ]inpL'r thu anthnr ropapitnliiti'd -with ad<liti()ns thodi'.«(ii|iti(in of tlir l<i(es 

 and a])])aratn.s eni])l()y('d, tiigi'tlur with the nietliod of nmliiii;.' tho observalions 

 rthvady given in tlio ' (^nartcrly Jonrniil' of the .Moteorolo^'-ii'al Sociuty in ]ns;{, 

 'I'wd l;itoa are ilown tandem, and tho lower or main one carries iihout l'.OOO feet of 

 piniio cord wire similar to tliat used by Sir AVilliam Thomson for his deep-sea 

 sounder. 



^\'hen tho observations are grouped roughly lor altitude only, the following 



results uro obtained for the exponent x in tho empirical furniula -■.-.( )' -wliero 



V, V, IT, /<, nro tho velocities in feet per minute and heiglit in leet at tho upper and 

 lower elevationa respectively. 



Xii. of 

 Obscrvatidns 



Moan 



upper lieiylit 



iibiive )iliii'(! 



of ol)si'rvati()ii 



in feet 



Thus, while the velocity is invariably found to increase, the rate of increase 

 ra])idly diminishes above th(> tirst I'OO or .')00 feet. Xear the sui-faco the increase 

 of velocity is probably very rapid, and possibly agrees with Mr. Stevenson's lornuila 



- . Above the tirst fifty feet, bow ever, it is plain that this formula will not 

 /' n 



hold. It must further be remembered that the station of observation is itself 

 500 feet above sea-level, and that about 200 feet above the ground tho motion of 

 the air corresponds nearly with what it would have at its real height above a 

 sea-levid plain. On Ibis assumption, and adding the fiOO feet lo both the I'levations 

 in the two last groups of the table, we get .r^S and ! res])eetively, valu(>s which 

 are probably nearer the truth than 3 and j^,., uikI which hover round the mean 

 value :|^, determined by tho author from a discussion of cloud velocities up to a 

 height of 2.'>,()00 feet above sea-level given by Dr. Vettin of Bi rlin, and discussed 

 by the author in ' Nature ' for May 188.3. 



Amongst other indirect results obtained by tbe author is the fact which tends 

 strongly to support ])r. Kijppen's theory of the diurnal period in the velocity of the 

 surface-wind Tbe aiithor has frequently found that during the day his Idte tlew 

 with dilliculty, while in the evening it invarialjly tlew steadier, higher, and with a 

 stronger pull than during the day. According to Dr. Kcippen there is an inter- 

 change of air (Luftanstausch) between the upper and lower layers occasioned by 

 the surface heating during the day, which tends to increase tho velocity of the 

 lower air by tbe communication to it of tho more rapid motion which the descend- 

 ing air brings with it from above. ]\v an exactly converse act i(m the motion of 

 the upiier layers is correspondingly diminished l)y the retarding action of the 

 ascending air. This theory agrees perfectly with the phenomena observed by the 

 author. ■*• 



Evidence as to the existence of an ascending currtmt under the front of cumuli 

 and cumulo-strata, and a descending current under their rear portions is also fur- 

 nished by the behaviour of the kites. 



' Printed in extcnso in Nature, November 20, 18S1. 





