61 
1887.] Mr John Aitken on Thermometer Screens. 
tions were made, or it might have resulted from the hour at which 
the high maximum happened to be reached, or it might even have 
been produced by an error in our time-keeper, either advancing or 
retarding the time of observations, while keeping them at hourly 
intervals. It is evident, therefore, that curves of temperature drawn 
from readings taken at intervals, unless very short, have hut little 
value, and may be most misleading. It would he better therefore, 
instead of curving the results, simply to connect them with straight 
lines, so as to enable the eye easily to catch the successive readings, 
it being understood that these lines give no information as to the 
state of matters between the observations. 
New Screen. 
In Part III. of this communication ( Proc . Roy. Soc. Eclin ., 
Ho. 121) reference was made to some attempts made to check the 
entrance of radiant heat through the air passages into the draught 
tube screen. The most successful results were got by introducing 
small screens between the bulb of the thermometer and the source 
of radiation, and so arranging the air circulation that all air that 
had touched the radiation-heated surfaces was drawn away through 
side passages, and only the central core of unheated air allowed 
to pass on to the centre of the screen, and come into contact with 
the bulb and its surroundings. The plan, however, which was 
found to work well in a room, in still air, was found to be quite 
unsuitable for observations in the open air, owing to the wind 
causing eddies inside the screen which interfered with the proper 
circulation of the air, and mixed the air heated on the screens with 
that entering the centre chamber ; an attempt was therefore made 
to see if the principle could not be modified to enable it to be 
applied in a form suitable for open-air observations. 
A number of complicated forms suggested themselves, but as all 
of them seemed likely to give rise to eddies inside the screen, they 
were abandoned without trial. At last the simple form shown in 
fig. 1 was designed, and a screen of this form was constructed 
at the beginning of October. The figure shows a vertical sec- 
tion through the centre. As will be seen, it is of extremely 
simple construction. The circulation of air through this screen is 
entirely affected by the natural movements of the air. Ho modifi- 
