236 Proceedings of Royal Society of Edinburgh. [sess. i 
the cyclonic, then this at first sight might seem to indicate that 
anticy clonic air is more impure than cyclonic; that is, that air 
moving from areas of high pressure, which is generally supposed to 
come from the upper strata, has more dust than that near the earth, 
which would be equivalent to saying that meteors falling into our 
atmosphere produce more dust in the upper strata than is thrown 
into the lower by all the sources of pollution on the earth’s surface 
in country districts. 
There is, however, evidently another interpretation of these facts. 
One marked difference between a cyclonic and an anticyclonic area 
is, that the isobars in the former are much closer together than in 
the latter; that is, the barometric gradient is steeper in the cyclonic 
areas; or what this indicates is that there is more wind in the 
cyclonic than in the anticyclonie areas. That is to say, that in cyclonic 
areas there are high winds and little dust, while in anticyclonic 
areas there is little or no wind, and the amount of dust is great. 
Accepting this explanation of the greater or less amount of dust 
in the two areas, let us see how it is borne out by the observations. 
In order to investigate this point we may lay off a series of 
points representing the distance between the isobars, on the 
different dates. The height of this curve would be inversely pro- 
portional to the velocity of the wind. Instead of using the baro- 
metric curves, we will use the observations of the velocity of the 
wind observed at the stations. These observations are entered in 
the diagram and connected by a fine line. It will be seen in 
diagram Ho. 1 that the fall in dust which took place on the 6th 
was accompanied by a rise of wind, and that the fall of the wind on 
the 9th was accompanied by an increase in the dust. The wind 
increased on the 10th, and the dust fell; but on the 12th, 13th, 15th, 
and 16th both curves rose and fell together. In diagram No. 2 it 
will be seen that both wind and dust rose and fell together from 
the 5th to the 9th, but on the 9th the wind increased and cleared 
away the high dust that had prevailed for some days. From the 
9th to the 13th the dust increased whenever the wind got less, 
but on the 16th a great increase in dust took place with a slight 
increase in wind. When, however, we come to diagram No. 3, the 
relation between the wind and the dust is much more marked. 
It will be observed that almost every rise of wind is accompanied 
