PRESENT STATE OF OUR KNOWLEDGE OF THE UPPER ATMOSPHERE. 115 
which the ascents in these months were made, was accidentally peculiar 
for the five years of observation. 
The following table gives the mean pressure at Strassburg for each 
month on the days of ascents during the five years, and for comparison the 
mean pressure at Aachen ' from the records for 1896-1903 :— 
Mean Pressure (700+ mm.). 
— Jan. Feb. Mar.) Apr.| May June July Aug. Sep. Oct. Nov.) Dec. 
ded | | | | 
pooh 4: bvealh .-iec | 
Strassburg (140 m.) . | 54 | 55 | 49 | 49 | 53 
Aachen (169 m.) . .| 49 | 47 | 44 | 46 | 46 
51 | 61 | 49 | 50 | 50 
47 | 47 | 47 | 47 (46 48 
48 | 
45 | 
The differences in pressure are practically no greater than are to be 
accounted for by the ordinary yearly variation. Of course the same 
causes which affected the value of H, and the mean temperature of the 
upper air would also affect the pressure. The possibility of such accidental 
peculiarities, however, emphasises the need for weekly ascents at one or 
two selected stations to fill up the large gaps between successive inter- 
national ascents. , 
IV. (e) Diurnal Variation of Temperature. 
Clayton concluded from a discussion of kite ascents made at Blue 
Hill that the most marked feature in the diurnal temperature variation 
in the free atmosphere was the increase in the semi-diurnal term, and 
the vanishing and reappearance with changed phase of the diurnal term 
in the first 1000 m. Wundt,? using observations made at Hald, in Jutland, 
obtained results for 1,200 m. from which the following expression for the 
variation has been calculated— 
T=T,+0'55 sin (nt + 248°) + 0:05 sin (2nt + 349°), 
time being measured from midnight and the amplitudes being expressed 
in degrees C. 
For the autumn of 1902, his results, which he regards as more trust- 
worthy than the general means, give 
T=T, +0:35 sin (nt + 229°) + 0°13 sin (2Qnt + 217°). 
Gold? found the following values for 1,000 m. from an analysis of 
the kite and balloon observations made over Berlin and Lindenberg in 
1903-07 :— 
'! = T,— (4°40 + 0°08) + (0°87 +£.0°13) sin (nt + 197°) 
+(0:14+0:10) sin (2nt + 123°). 
where T, is the mean surface temperature at Potsdam (40 m.). This 
agrees with Wundt’s results in making the amplitude of the semi-diurnal 
term small, but differs from them considerably in the amplitude of the 
diurnal term and in the phases of both terms. 
A similar result was found when only those observations were used 
in which the observed wind at 1,000 m. was not less than 8 m.p.s. 
' Met Zeit., 1906, p. 91. 2 Thid., 1908. 
3 Nature, July 1, 1909. \ 
