I'ISTKIBUTION IN THK ATMOSPHERE OVER ENGLAND. 207 



It will be seen from these tables that the average maximum height reached is 

 greater in summer than in \vinter. This is partly due to the fact that the bursting- 

 point of a balloon depends on the pressure, and in summer a greater height must be 

 reached to attain the same low pressure. This does not account for the whole 

 variation, and the part left must be ascribed to the effect of temperature on india- 

 rubber. It becomes hard if left unused hi cold weather. For reaching great heights 

 the quality of the rubber is of supreme importance. 



The figures in column B are instructive. Large inversions of temperature are to be 

 met with at the commencement of the isothermal in summer rather than in winter ; 

 and at times when the barometer is near its normal value, and not when it is low. 



The figures in column C are not what would be expected. The surface wind in 

 England is certainly stronger in winter than in summer, but the distances run are 

 reversed and are rather greater in summer than in winter. The exact time taken for 

 any ascent is not known, but judging from the few cases where the balloon has been 

 seen to fall it may be assumed to be about two hours. 



In the pressure group, places with an average barometer are naturally places where 

 there is a barometric gradient, and hence a long run for a balloon is to be expected, 

 but the value 36 miles, the lowest in the table, for the six cases that occurred in or 

 near the centre of a deep cyclone does not favour the idea that cyclones are eddies 

 produced by a strong upper current in the general circulation. So many of the 

 balloons have doxibtless fallen in the sea that there must be large systematic errors 

 in the means of the distance run. Balloons often fall within a short distance of their 

 starting place, and it is certain that there is no strong and permanent upper drift 

 from the west. 



With reference to the bearing of the place of fall, no definite statement can be 

 made. The prevailing wind in England is from some westerly point, and so in general 

 a balloon falls towards the east ; they seldom seem to go due west, but south-west 

 and north-west are quite common directions, and I do not think the frequencies of 

 winds from various points of the compass are very different above and below. On 

 many occasions a strong current from the north and north-west has l>een met with at 

 great heights when it has been calm up to the cirrus level, and while the track of a 

 balloon will in general follow an isobar, the track cannot be foretold with certainty 

 from the surface distribution of pressure. 



Wave Motion on the Lower Boundary of the Isothennal Surface. 



In a fair number of the ascents that are made it is found that the values of H c and 

 of T,; have changed in the short time that the instrument has remained in the 

 isothermal region. The variations are much too large to be ascribed to casual 

 instrumental errors, neither are they due to any systematic lag of the aneroid or 

 of the thermograph, for they will sometimes be shown on one instrument in one 



2 M 2 



