The relative humiditj- of the atmosphere was usually very small, the 

 difference between the wet and dry bulbs generally exceeding ten degrees, and 

 not infrequently exceeding fifteen degrees. 



Jwne. 1909. Lan-chou Fit. Alt. 5106 Ji. 



June was decidedly hotter than May, and the rainfall was greater. The 

 prevailing wind was no longer in the east, but blew from the north-east. 



On the gth, at 3.30 p.m., a fierce gale swept up from the west, tearing 

 branches off the trees and carrj'ing them for considerable distances. No 

 serious damage was done, however. A strong wind and a dust storm were 

 recorded on the loth ; a strong wind on the nth ; and another on the 12th. 

 During these three days of disturbance the barometer fell from 24*90 inches 

 to 24-68 inches, rising once more to 25.06 inches. On the 12th rain fell, and 

 continued intermittently till the 17th. Heavy rain was also reported by two 

 members of the party at work in the mountains to the south of Lan-chou on 

 the i2th and 13th ; whilst a heavy deluge, preceded by a violent wind, was 

 recorded by one of them on the 14th. Rain was again recorded on the 15th 

 in the same district, while a little to the east of this position heavy rain was 

 recorded on the 20th. On the 22nd a west wind was noted, followed by 

 rain, and again on the 26th a west wind brought up rain clouds, when 72 of 

 an inch of rain was recorded in sixteen hours. The wind continued to blow 

 from the west for two days, then, shifting right round to the east, seemed to 

 bring back the rain clouds. 



It must here be noted that rain was frequently brought up by a north- 

 easterly wind. The fact that the Yellow River flows in a north-easterly direction 

 from Lan-chou Fu may account for this. It cannot be doubted that the 

 Kokonor has some effect upon this district, but it is equally certain that it is 

 not the only agency at work, and an analysis of the wind direction and 

 corresponding periods of rain leads to the idea that the large volume of water 

 flowing down the bed of the Yellow River is another powerful element in the 

 meteorology' of the district. Only on very rare occasions (in April) was rain 

 recorded from any quarter but north-east and west. The total rainfall for the 

 month was i"i4 inches, notwithstanding the fact that on five occasions 

 the fall was not measured. 



Clouds appeared every day of the month. Cumulus formations pre- 

 dominated, though stratus clouds were not infrequent, cirrus clouds occurred 

 more often than in May. 



The maximum temperature was 90*0° F. on the 26th and 29th. The 



