OCTOBKR, IQll. 



KNOWLEDGE. 



377 



temperature decreases rapidly with height, namely, 

 at the rate of 1°F. for every three hundred feet. 

 It is now well established that, in the lower two 

 miles, this is frequently reversed, a fact jireviously 

 thought exceptional. But from this height up, as 

 shown by the closely-set parallel lines of the model, 

 there is great and constant regularit\-. Formerly, it 

 was supposed that this continued unbroken until the 

 cold of space was reached, presumably somewhere 

 near absolute zero, or — 273"C. No one dreamt of 

 an\ serious di\'ergence. When, however, the 

 bdlloiis sondes records began to accumulate, one 

 of the most obvious facts obtained was that at a 

 hcMght, usualh' of about si.x miles, temperature 

 cl'cisl'c/ t(i fall, and on the contrary tended again 

 to rise, and that this unlooked-for reversal, so 

 extraordinar\- that as yet we have for it no 

 certain explanation, continues as high as we 

 ha\e at present an^" observations, namely, at 

 the least up to a height of fifteen miles. 



In other words, whilst 

 we still cannot but believe 

 that, outside our atmos- 

 phere, the cold of space 

 is intense beyond easy con- 

 ception, so far as actual 

 knowledge goes, it ceases 

 to grow colder at a height 

 of six miles and remains 

 prac t ical 1\' u nal tered 

 tliiDiigii, at any rate, the 

 succeedmg eiglit or nine 

 miles. 



This 1 a %■ e r was first 

 named, verv appropriately, 

 the " isothermal layer," 

 but. to connect it with 

 the regions of change 

 beneath, these latter have 

 been called the " tropo- 

 sphere," the upper the 

 " stratosjihere." The latter 

 lies abo\'e the highest \isible 

 clouds. 



The average height of 

 the lower surface of the 

 stratosphere, which is. how- 

 ever, subject to consider- 

 able local perturbation, 

 alters both w ith the season 

 and latitude, being, in 

 Europe, least (9-1 kilo- 

 metres, or about tive-and- 

 a-half miles) in March, and 

 highest (11-9 kilometres or 

 about seven miles) in 

 October. The only decided 

 break in the curve is for 

 September, a kilometre 

 lower than August, and 

 one-and-a-half kilometres 

 below October. It is 



that here also occurs, at least in the 

 British Isles, the chief break in the rainfall curve. 

 This, also, is the time when the north trade 

 winds are at a minimum, and there may well be 

 some association. The temperature is lowest in 

 February ( — 80°F.), highest in September ( — 60°F.). 

 Observations outside European latitudes are still 

 exceedingly rare, but some special ascents in East 

 Africa, over the Atlantic, and in the West Indies, 

 have given important evidence as to equatorial con- 

 ditions. Over the Atlantic the stratosphere was not 

 reached at fifteen kilometres, which is nearly nine 

 miles, or half as high again as the European average. 

 It was, however, revealed by two ascents made on 

 the Victoria Nyanza, which reached a mile and three 

 miles higher (ten and twelve miles). Here, in 

 equatorial regions, was recorded the lowest air 

 temperature \et obtained, 119"F. below zero 

 ( — 84'-''C). Thus we have the interesting result 

 that the regular fall of temperature continues up 



CURVES SHOWING CHANGE OF TEMPERATURE WITH HEIGHT ABOVE SEA- LEVEL 

 OBT/^NED FROM BALLON- SONDE ASCENTS 1907-8. 



-40 -20 IQ 



TEMPERATURE IN DEGREES TAHRENHEIT , 



230 Z40 2^0 ?60 270 



TEMPERATURE ABSOLUTE 



/Us I^Inffsty's Staiifltiery Office. 



Figure 6. Relation of Temperature and Height. 



300 



