1887 .] Mr John Aitken on Thermometer Screens. 
67 
following is called screen A, and the old one B. The screens were 
fitted upon the lawn at a distance of about 15 feet apart, in as nearly 
as possible similar exposures to sun and wind. This was done on 
the 14th September, and in order to make sure that both screens 
gave the same temperature, comparative trials were made with 
them, the bottams of both being open. Though the day was not a 
very suitable one for the test, as there was not much radiation, still 
the old screen was found to give higher readings than the other 
by about half a degree. This was owing to the paint being dirty, 
and the surface of its louvres being better absorbers of beat than 
the louvres of the new and clean one. The screens were now painted 
white, after which they looked nearly alike in whiteness, and on trial 
with both bottoms open were found to give readings nearly alike. 
As it would be almost impossible to set up the two screens in posi- 
tions exactly alike with regard to exposure to wind, radiation, &c., 
both screens were provided with movable bottoms, so that either 
could be worked with the bottom closed while the other was open. 
When testing the screens, sometimes the same screen was kept closed 
throughout the whole day, at other times first the one then the 
other would be closed, while the external conditions remained con- 
stant, so as to check any difference in temperature due to position 
or condition of screen, direction of wind, &c. 
In these trials the ordinary thermometer used for taking wet and dry 
bulb observations was employed. The wet bulb with its apparatus 
was removed, and the thermometer in each screen was placed where 
the wet one usually is, the index of the instrument being turned 
slightly round, so that it could be read by opening the door to only 
a very small extent. The object of this was to prevent radiant beat 
entering and altering the readings while they were being taken; also 
to prevent radiation beating the inside of the screen. The thermo- 
meters with which most of the observations were made bad round 
bulbs about 8 mm. diameter, but others with smaller bulbs were 
occasionally used. All the thermometers were graduated on the 
stems, had wide scales, and were easily read. In addition to the 
Kew corrections, they were all carefully compared with each other 
in water at intervals of 5 degrees or less. The room where these 
comparisons were made was heated to the temperature of the 
water, so that the temperature of the large volume of water surround- 
