632 
Proceedings of the Royal Society 
Added June 23. — Further experiments made with larger elec- 
trodes show that the above empirical formula closely represents the 
results of observation; but they also show that the curves are 
symmetrical about an axis, and are probably hyperbolas. I am at 
present engaged in investigating the point. 
6. On Thermometer Screens. By John Aitken, Esq. 
(Plate XXIII.) 
Part III. 
In my previous communications to the Society on this subject * 
some different forms of thermometer screens are described, and the 
results are given of a number of trials made with them during trying 
conditions of weather. The result of these tests showed that all the 
screens gave readings a little too high when there was much radia- 
tion, and that the thermometer in the fan apparatus also indicated 
too high a temperature, the readings of a fine-bulbed thermometer 
in a polished silver case being taken as our standard of temperature 
in these trials. The reasons are given for the thermometer in the 
draught screens indicating too high temperature, and the ineffectual 
attempts made at the time to check these errors are described. 
Confining our attention for a moment to the draught screen, fig. 
3, and the fan apparatus, figs. 5 and 6, of previous papers. The 
reasons for the thermometer in screen fig. 3 reading too high are — 
1st, the outside of the case being heated by the sun, part of this 
heat is conveyed inwards to the thermometer, being conducted 
through the concentric tubes, and radiated from tube to tube, 
the air passing over the surface of the tubes not being able complete- 
ly to check this transference of heat ; 2nd, the too high readings 
given by this screen are also in part due to the heat radiated 
and reflected from the ground, and entering the lower end of the 
tube. The introduction of small screens into the tube below the 
thermometer to check the entrance of radiant heat, while admitting 
a free passage of the air, was not found to improve the correctness 
of the readings given by this screen. The entering air seemed to be 
heated as much by the screens as the bulb was by direct radiation. 
* Proc. Hoy. Soc. Edin., No. 117, 1883-84. 
