312 
pared with frost-point hygrometer measurements made 
at the same time, they should provide a useful check 
on the infrared absorption data. The method is being 
developed by G. B. B. M. Sutherland and R. M. Goody 
at Cambridge, England. 
Simon [16] has suggested that enough air to allow 
accurate analysis could be condensed in a small con- 
tainer cooled in liquid hydrogen and carried by a bal- 
loon to the height at which the sample is to be col- 
lected. This method is now being developed by F. E. 
Simon and A. J. Peckover at Oxford. 
The dew- or frost-point hygrometer has been found 
satisfactory even at low temperatures. It has the great 
practical advantage that it needs only a very simple 
calibration, since all that is required is the temperature 
of the thimble when a thin deposit of hoarfrost is neither 
increasing nor decreasing. At —80C it is necessary to 
detect deposit changes of the order of 10° grams of 
ice, corresponding to a uniform deposit on the thimble 
surface of the order of thickness of one atomic diam- 
eter. This is not impossible because the ice is deposited 
as a large number of small crystals which scatter light 
very efficiently, so that changes in the deposit can be 
detected by changes in the light which it scatters. This 
will be understood, since only 10—” grams is equivalent 
to 1000 crystals each 0.4 » across. With proper instru- 
ment design the change in the deposit can be detected 
by eye, though the observations are difficult when the 
frost points to be measured are below about 200 to 
210K. A photoelectric method of estimating the amount 
of light scattered by the deposit, though more compli- 
cated, is usually preferable. 
DESIGN OF A FROST-POINT HYGROMETER 
Three types of instruments have been employed in 
Great Britain: (1) an instrument in which the deposit 
is observed visually, and the cooling of the thimble 
performed manually, (2) an instrument in which the 
amount of deposit is indicated photoelectrically, and 
the cooling operated by hand, and (8) an automatic 
instrument in which the thimble is held constant at the 
frost point. 
If the instrument is to measure the frost point of 
very dry air accurately (e.g., frost points down to 190K), 
there are certain conditions which must be observed: 
1. Air Supply. A constant source of trouble, when 
measuring very dry air, is the possibility that the air 
might pick up water vapour from the walls of piping 
leading to the hygrometer. In aircraft observations this 
error is guarded against by passing a large and rapid 
current of air through the tubing leading to the hy- 
grometer. Most of this air passes out of the aircraft 
again and only a small fraction is used in the hygrom- 
eter. The branch pipe actually leading to the thimble 
is kept very short. 
2. Determination of the Amount of Deposit. As stated 
above, the only way in which it has been found possi- 
ble to observe the very small deposit which is obtained 
with very dry air is to illuminate the surface of the 
thimble strongly and measure the light scattered by 
the deposit. The surface of the thimble must be highly 
THE UPPER ATMOSPHERE 
polished so that little light is scattered by it. If the de- 
posit is detected by a photoelectric cell, the light scat- 
tered into the cell by the clean thimble must be kept 
as small as possible, so that the additional light scat- 
tered by a faint deposit is as large a fraction as possi- 
ble of the total light entering the photocell. 
3. Temperature Control of the Thimble. For the hand- 
operated hygrometer the thimble has been cooled by 
cold petrol (cooled in a surrounding bath of petrol and 
solid CO), or by liquid air, forced as a jet against the 
hollow underside of the thimble. With suitable design 
it is easy to control the temperature of the thimble by 
varying the rate of pumping. An alternative—which is 
always used in the automatic hygrometer—is to cool 
the thimble by conduction to liquid air and to warm it 
by an adjustable heating current. 
GLASS ELLIPSE 
ILLUMINATING 
LAMP 
S 
HEATER 
RESISTANCE ~ RSS SSS 
THERMOMETER . 
WINDING Surat: 
OPERATING 
HANDLE 
DEWAR VESSEL —— 
Fic. 1—The frost-point hygrometer. A deposit of dew or 
hoarfrost on the top surface of the thimble is illuminated ob- 
liquely and observed by eye through a X7 magnification. The 
thimble temperature is controlled by operation of the pump 
and measured by a platinum resistance wire wound on the 
thimble skirt. (Reproduced by courtesy of the Royal Society, 
London.) 
Tn the automatic hygrometer the scattered light from 
the thimble is received by a photocell or photomulti- 
plier. A second photocell or multiplier receives a con- 
stant small light from the same source which illuminates 
the deposit, and the difference of the photocurrents for 
the two multipliers is amplified. To avoid hunting, a 
differentiating circuit is usually necessary so that the 
final control of the heating current depends both on 
the amount of the deposit and on its rate of growth 
(or evaporation). 
One of the difficulties with which an automatic hy- 
grometer must contend is the very extended range of 
the absolute humidity of the air which it must measure. 
Also, the rate of growth of ice crystals decreases pro- 
gressively at temperatures below about 218K, and for 
practical purposes crystals do not grow at temperatures 
