PRESENT STATE OF OUR KNOWLEDGE OF THE UPPER ATMOSPHERE. 83 
and the zero trace are marked side by side on the usual smoked aluminium 
sheet fixed to a revolving drum. The scale of the instrument is :— 
1 mm. mercury = 0°1 mm. deflection. 
1°C. == 0°7 mm. deflection. 
The clock is of invar and is guaranteed not to stop even at —80° C. 
The instrument is provided with a protecting case and weighs 750 gm. 
There is no forced ventilation, the rate of rise and fall of the balloon being 
deemed sufficient protection against solar radiation. In an ascent the 
instrument is suspended by springs in a basket lined at the sides with 
nickel paper. 
In order to reduce the weight of the instrument for use with small 
tubber balloons Assmann abandoned the heavy clockwork, and, after 
various modifications, devised the following instrument. Two cylinders 
free to rotate and with their axes parallel are mounted one above the other 
in the frame, and the record sheet forms an endless belt round them. One 
of the cylinders is turned on its axis by the expansion or contraction of the 
multicellular aneroid barometer, and the other cylinder and the record 
sheet move with it. The thermometer pen is carried across the sheet, 
parallel to the axes of the cylinders, by an endless thread passing round 
two pulleys, which are caused to turn by levers connected with a bimetallic 
thermometer consisting of copper and invar strips soldered together. 
The pressure is thus indicated by the movement of the record sheet, 
and the temperature by the movement of the pen across the sheet, the two 
motions being exactly at right angles to each other. The humidity is 
indicated in the same way as the temperature, a double-span hair hygro- 
meter being used. A small clock draws a pen across the record sheet to 
indicate the duration of the ascent, and to show if the balloon burst 
instantaneously on reaching the maximum height. Stoppage of the clock 
does not materially affect the results. The thermometer strip and the 
hygrometer hairs are enclosed in a ventilation tube, and in some of the 
instruments are aspirated by means of an electrically driven ‘ Scirocco ’ 
fan fixed into the ends of the ventilationtube. The total weight of instru- 
ment and case is 620 gm. 
The Dines meteorograph is of quite different design.' It is a baro- 
thermograph, no measurements of humidity being attempted. The 
barometer 1s in general a partially exhausted German-silver aneroid, and the 
thermometer is bimetallic, consisting of a strip of aluminium or German 
silver and arod ofinvar. The partially exhausted aneroid is used because 
it gives a larger scale than the totally exhausted box.? In the Assmann 
instrument the record sheet moves bodily, while the barometer and 
thermometer elements are fixed; in the Dines instrument the same 
effect is obtained by making the barometer and the record sheet fixed, 
while the thermometer moves bodily. The aneroid is fixed on one side to 
the frame of the instrument, and on its other side carries the thermometer. 
When the aneroid expands or contracts the thermometer is moved laterally 
asa whole, and the two pens, being attached tothethermometer, are carried 
across the record sheet and mark two similar and parallel pressure traces. 
The German-silver strip and the invar rod of the thermometer are 
approximately of equal length, straight and parallel to each other, and 
are separated slightly. They are fixed together at one end, and the 
' Symons, Met. Mag., July 1906, p. 101. 
* The original and perhaps more important reason was to utilise, as far as 
possible, the more perfect elasticity of a gas instead of that of a metal. The boxes 
containing air have very little lag. 
G2 
