of Edinburgh, Session 1888-84. 
667 
of air it said the humidity was only 67. From these observations 
we see the necessity of keeping the wet bulb at a distance from all 
objects which interfere with the free circulation of the air. But 
even with the very best form of construction,^ the readings taken 
at different times, and by different instruments, can not be com- 
pared with each other unless we can control the circulation. 
For the purpose of getting the true temperature of the air with 
the dry bulb, the Froude, or sling thermometer has been devised, 
but its evident inconvenience and risk have prevented its general 
use. An improved form of this method of observing has been 
designed by Mr Edwin Clark. In his observations he uses a whirl- 
ing table, to which the instrument is attached, the table with the 
thermometer being rotated rapidly before the readings are taken. 
Mr Clark has also used his whirling table for wet bulb observations. 
By these means a decided advance has been made, and a nearer 
approach to correct readings of wet and dry bulb thermometers has 
been obtained. The practical difficulty in working this apparatus 
is the impossibility of taking the reading while the air is acting on 
the instruments ; and as the table requires to be stopped before the 
readings can be taken, there is time for them to change. To obviate 
this difficulty, I have arranged for wet and dry bulb observations 
a simple revolving fan arrangement for keeping up a circulation of 
air past the thermometers, which brings them quickly to the correct 
temperature, and allows the readings to be made while the air is 
rushing over the instruments. 
This apparatus is shown in figs. 5 and 6. . Fig. 5 is a back eleva- 
tion, and fig. 6 a side elevation shown in section. The bulbs of 
the thermometers t, t, are placed in the horizontal tube a ; a vertical 
division being fixed in the tube between the bulbs. The tube a is 
placed at a height of 4 feet from the ground, and is supported 
by the two standards hh, which are firmly fixed in the ground, or 
otherwise supported. For circulating the air a large fan is used, 
as it enables us to dispense with all multiplying gear, which is 
troublesome, and aj)t to get out of order when exposed to the 
weather. This fan is shown at c ; it is driven by the handle d, 
which has a small radius to enable us to turn it rapidly.* The 
* The handle has been coupled by means of a connecting-rod g, with a short 
lever or treadle /, if we may use the word, which is actuated by the hand. This 
