of Edinhurgh, ScssioJi 1883-84. 
681 
centre tube was found to be generally about half a degree higher 
than in the centre tube, while the outer passage was often more 
than a degree higher, showing that the lower end of the tube is not 
properly protected. Another screen is therefore being prepared, with 
a more perfect screen protection, a larger draught tube, and more 
concentric non-conducting tubes and air passages, and it is hoped 
this new screen will have a less error. It will be remembered that 
the fan thermometer is, under the conditions during the experiments, 
always about one degree below the sling thermometer, so that screen 
fig. 3 is nearly half a degree nearer the truth than the sling. 
It is only within the last few days that I have been in possession 
of ■ a Stevenson screen, and been able to make comparative trials 
with it. During these tests its action was compared with the 
fan apparatus, and readings were taken at the same time of the 
thermometers in the draught tube screens. These latter, as already 
stated, had an average error of about 0°*6 too high. The smallest 
error recorded in more than thirty readings of the thermometer in 
the Stevenson screen was 1°’3, and it only fell to that on two 
occasions. The excess error was generally more than 2°, and was 
as high as 2*8 on two occasions. The morning on which the trials 
were made was certainly a trying one, but was not so bad as it 
might have been, for the Stevenson screen. The sun was very 
strong, and there was but little air moving ; the direction of the 
wind, however, was favourable, being from the north-east, so that 
the air entering the screen passed over the cold louvre boards and 
not over the sun-heated ones. If the wind had been southerly, 
the error would have been greater, as the entering air would have 
been heated in its passage over the hot louvres. These tests were 
made on an exceptional day. On most days, when there is wind, the 
screen gives a much nearer approach to the truth than it did on that 
occasion. By a slight addition to this screen, I was able, when the 
error was high, greatly to reduce it. On those days on which the 
wind blew from a northerly direction, and the air entering the screen 
came in from the cool side of the box, I found the high temperature 
given by the thermometer was due to radiant heat entering the open 
bottom of the box. Either a solid or a louvred bottom has been 
found greatly to improve the correctness of this screen, but whether 
the solid or the louvred bottom is the best has not yet been deter- 
