TRANSACTIONS OF SECTION A. 



G39 



(half an hour) required for polishing when little alteration of figure be required, 

 and -when no scratches have to be worked out. 



13. An Account of some preliminary Experiments with Biram's Anemometers 

 attached to Kite-wires. 1 By Professor E. Douglas Archibald, M.A. 



In this paper the author recapitulated with additions the description of the kites 

 and apparatus employed, together with the method of making the observations 

 already given in the ' Quarterly Journal ' of the Meteorological Society in 1883, 

 Two kites are flown tandem, and the lower or main one carries about 2,000 feet of 

 piano cord wire similar to that used by Sir "William Thomson for his deep-sea 

 sounder. 



"When the observations are grouped roughly for altitude only, the following- 

 results are obtained for the exponent x in the empirical formula — = ( — V where 



V,- v, H, /*, are the velocities in feet per minute and height in feet at the upper and 

 lower elevations respectively. 



No. of 

 Observations 



Mean 



upper height 



above place 



of observation 



in feet 



249 

 412 

 634 



Mean 



lower height 



in feet 



93 



173 

 324 



Mean upper 

 velocity in 

 feet per nrin. 



1.630 

 1,751 

 1,987 



Mean lower 



velocity in 



feet per min. 



1.145 

 1,474 

 1,902 



Approximate 

 value of x in 



the formula 

 V = /H\x 



v [h' 



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

 rapidly diminishes above the first 200 or 300 feet. Near the surface the increase 

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



— = — . Above the first fifty feet, however, it is plain that this formula will not 

 v h 



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 the motion of 



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



sea-level plain. On this assumption, and adding the 500 feet to both the elevations 



in the two last groups of the table, we get x = i and 4 respectively, values which 



are probably nearer the truth than § and -^, and which hover round the mean 



value ^, determined by the author from a discussion of cloud velocities up to a 



height of 25,000 feet above sea-level given by Dr. Vettin of Berlin, and discussed 



by the author in ' Nature ' for May 1883. 



Amongst other indirect residts obtained by the author is the fact which tends 

 strongly to support Dr. Koppen's theory of the diurnal period in the velocity of the 

 surface-wind The author has frequently found that during the day his kite flew 

 with difficulty, while in the evening it invariably flew steadier, higher, and with a 

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

 change of air (Luftaustausch) between the \ipper and lower layers occasioned by 

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

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

 ing air brings with it from above. By an exactly converse action the motion of 

 the upper layers is correspondingly diminished by 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 current 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. 



1 Printed in extcnso in Nature, November 20, 18S4. 



