310 



enable the thermometers to assume with more rapidity the tem- 

 perature of the surrounding air, and also to diminish the effect of 

 radiation, in case the shades should not be a sufficient protection, 

 especially when the balloon was stationary or rising very slowly. 

 The thermometers used were very delicate, the bulbs being cylinders 

 about half an inch long and not more than y^th of an inch diameter. 

 It was found on trial that when the bulbs were heated 20° above the 

 temperature of the air in a room, they resumed their original reading 

 in 40 or 45 seconds, when moved through the air at the rate of 5 or 

 6 feet in a second. It is thus probable that any error arising from 

 want of sensibility in the thermometers will be small, and in all like- 

 lihood not more than may be expected from other accidental causes. 



The observations were taken at short intervals during the ascent, 

 it having been seldom practicable to obtain a regular series in the 

 descent. The intervals were generally one minute, but frequently 

 only 30 seconds, so that an observation was for the most part 

 recorded every 200 or 300 feet. All the observations are given in 

 detail in the tables accompanying the paper. I'hey are also given 

 in the graphical form in the curves. 



The ascents took place on August 17, August 26, October 21, 

 and November 10, 1S52> from the Vauxhall Gardens, with Mr. 

 C. Green's large balloon. 



The principal results of the observations may be briefly stated as 

 follows : — 



Each of the four series of observations shows, that the progress 

 of the temperature is not regular at all heights, but that at a certain 

 height (varying on different days) the regular diminution becomes 

 arrested, and for the space of about 2000 feet the temperature 

 remains constant or even increases by a small amount : it afterv\''ards 

 resumes its dovrnward course, continuing for the most part to di- 

 minish regularly throughout the remainder of the height observed. 

 There is thus, in the curves representing the progression of tempe- 

 rature with height, an appearance of dislocation, always in the same 

 direction, but varying in amount from 7° to 12°. 



In the first tvv'o series, viz. Aug. 17 and 26, this peculiar inter- ' 

 ruption of the progress of temperature is strikingly coincident with 

 a large and rapid fall in the temperature of the dew-point. The same 

 is exhibited in a less marked manner on Nov. 10. On Oct. 21 a 

 dense cloud existed at a height of about 3000 feet ; the temperature 

 decreased uniformly from the earth up to the lower surface of the 

 cloud, when a slight rise commenced, the rise continuing through 

 the cloud and to about 600 feet above its upper surface, when the 

 regular descending progression was resumed. At a short distance 

 above the cloud the dew-point fell considerably, but the rate of dimi- 

 nution of temperature does not appear to have been affected in this 

 instance in the same manner as in the other series ; the phenomenon 

 so strikingly shown in the other three cases being perhaps modified 

 by the existence of moisture in a condensed or vesicular form. 



It would appear on the whole that about the principal plane of 



