57 
1887.] Mr John Aitken on Thermometer Screens. 
radiation. It will, however, he evident that its inertia is very- 
small, and w r hen exposed either without its silver sheath or when 
covered with chemically deposited silver, its sensitiveness is very 
remarkable, its indications showing constant fluctuations in the 
temperature of the air, as the mercury is in a continual state of 
pulsation whenever there is any radiation effect. These changes 
amount often to more than a degree in less than a minute, even in 
an October day, and are much greater in bright summer weather. 
The fluctuations do not seem to he due to variations in the radia- 
tion, as they are observed even when radiation appears to he con- 
stant, either under a clear sun or after it has been under a dense 
cloud for a time. These changes of temperature would seem to be 
due to the air that is heated on the ground and on other radiation 
heated bodies not being perfectly mixed with the colder air, one 
part of the eddy formed by the passing air having more heated air 
in it than another.* 
The curve FB, fig. 2, is drawn from temperature given by this 
fine-bulbed thermometer without any silver covering. The curve 
LB shows the readings of another thermometer with a larger bulb, 
exposed bare, alongside the fine-bulbed one. Its bulb is 22 mm. 
* This conclusion seems to be confirmed by observations made by Mr Dickson 
with this instrument at the top of Ben Nevis. He informs me that he never 
observed any of these rapid fluctuations at that station. The air heated on 
the slopes of the mountain will be carried away sideways by the wind, and the 
small amount of heating effected by the limited area of ground at the top does 
not seem to be sufficient to give rise to these changes. 
Professor Langley, in his celebrated researches with the bolometer, has 
observed certain fluctuations in the intensity of the solar radiation from 
minute to minute. As these fluctuations were ten times the instrumental 
errors, he is satisfied they have a real existence. He says the solar radiation 
would have been constant, but that the amount transmitted varied from minute 
to minute even in what appeared a cloudless sky. It is evident that the 
fluctuations observed by Professor Langley and those indicated by the fine- 
bulbed thermometer are not of the same order, although they take place at 
about the same intervals. The variations given by the thermometer are much 
too great to be due to variations in the amount of transmitted heat. As we have 
a satisfactory explanation of the fluctuations indicated by the thermometer, it 
seems possible that Professor Langley’s fluctuations may be due to the same 
cause. It is very difficult to suppose a want of uniformity in the upper air so 
great as to cause these fluctuations in such short intervals of time. It seems 
more probable that the variations observed with the bolometer are due to the 
imperfectly mixed hot and cold, moist and dry air near the surface of the 
earth, which will affect the readings of that instrument by absorption and 
by radiation. 
