ON THE SELF-REGISTERING ANEMOMETER. 323 
good. Had the anemometer been placed in a more exposed 
situation, we should have had the same comparative results, 
but they would probably have been more clearly developed. 
The principal thing to be aimed at in the selection of a site for 
such an instrument is, to have it equally exposed to all winds: 
whether it be on a hill or a plain, is of trifling importance, pro- 
vided it is not sheltered on either side*. 
There are many other facts which I had hoped to have adverted 
to, such as the law of succession, the mean duration of each 
wind, and the connection of the currents with the alterations in 
the state of the barometer, thermometer, and hygrometer, all 
of which must necessarily furnish interesting subjects for in- 
vestigation, though I have as yet been unable to devote to them 
a requisite portion of my time. 
* The anemometer from which these observations are taken being placed in 
the centre of a large town, does not afford such accurate results as would be 
obtained from an instrument situated beyond the reach of disturbing causes. 
Ses mC 
- ee 
y2 
