13G MESSES. W. N. SHAW AED M'. H. DINES: METEOEOLOGICAL OBSEEVATIOXS 
and 300U-3500 metres on the same two occasions with a fall of 1°‘7. There is 
considerable irregularity in the gradients for the first 500 metres. These irregu¬ 
larities may, perhaps, he accounted for partly by uncertainties attaching to the 
temperatures at the start and finish of the ascent when the instruments have to he 
handled hv the observer. The range of gradient was Ijetween 4° ft and 1° C. for the 
500 metres. The first case of gradient 4° (he., '8° for 100 metres) is on July G, 
Avhen an anti-cyclone was beginning to give way to an a})proaching de^^ression, and 
the gradient, although very high near the surface, disappeared altogether at ' 
1000 metres, where an inversion of gradient was disclosed with very dry atmosphere 
above a layer of clouds. Another occasion of similarly steep gradient under very 
similar barometric circumstances was on 24th July, but there was apparently no 
inversion of gradient. The depression in this case passed to the south of Crinan (see 
fig. 5) and gave north-easterly winds, while on the 9th it passed to the northward 
and gave westerly winds. The other occasion on which a steep gradient was 
oljserved in the lowest region presented no specific characteristics ; it may, however, 
l)e noticed tliat the temperature gradient up to 500 metres may be a good deal 
affected by the land in the neighbourhood which here and there reaches a height of 
some 1500 feet. 
Tlie very slight gradient of 1° C. for 500 metres on August 12 was connected with 
an inversion of tem})erature gradient not far from the surface. 
For the step Ijetween 500 and 1000 metres the fall of temperature reached 4° C. on 
two occasions, July 2G and August 1. Both occasions corre.sponded with the initial 
stages of advancing depi’essions. The forme]’ passed south of Crinan, the latter 
uorth. It seems that a general characteristic of the passage of a depression is that 
the isothermal lines on the diagram open out and a steep gradient becomes trans¬ 
formed into a slight one. As examples of tliis may be cited the transit ion between 
the 7th July and 9th July, which is better represented in the diagram than in the 
table; between morning and afternoon of the 26tli; between the bst August and the 
2nd; between the 19th A\igust and 20th, and between the 22nd and 23rd. 
Taking such corresponding heights for these dates as are available, there is an 
averao’e diminution of o-radient of as much as 50 iier cent., and the indication would 
seem to be that the columns of air in depressions are columns of relatively warm air. 
There are not enough observations at sufficient heights to confirm this suggestion. 
It is noticeable that in the depression the gradient does not reach that of the 
adiabatic gradient of saturated air in the higher regions. It will be noticed, on 
reference to the maps, that in every case the centres of the depressions were at some 
distance from the place of oliseiwation ; they ])assed it on all sides. 
In both months the temperature gradient fell off in the higher regions, and 
amounted to J'" for the stejis of 500 metres beyond 1500 metres. 
