80 REPORTS ON THE STATE OF SCIENCE. 
to verify barometer-height formule.' The ‘dromograph’ of Hermite 
and Besancon, a theodolite registering automatically the azimuths and 
angular altitudes of a balloon viewed from the ground, was also ex- 
hibited, as well as a heliometer employed by Kremser of Berlin for 
measuring the apparent diameter of balloons, and used since with pilot 
balloons. 
At this meeting it was resolved that— 
(1) Thermometers of less thermal inertia than those previously 
employed were necessary. 
(2) Efficient ventilation was indispensable. 
(3) Instruments should be tested before the ascents under circum- 
stances similar to those encountered during the ascents. 
(4) An aspiration psychrometer suspended at least five feet below 
the car was the only instrument suitable for manned ascents. 
At the third meeting of the Committee at Berlin, May 1902,? Assmann 
showed a compact apparatus for use with ballons-sondes, weighing only 
500 gm. The instrument registered through a pen filled with saltpetre 
on a sheet coated with lampblack which had been treated with a solution 
of ‘ Tonsol,’ and the resulting red trace could not be obliterated either by 
handling or by immersion in water. Hergesell and Teisserenc de Bort 
also exhibited self-recording thermometers. 
The type of instrument finally evolved for use with ballons-sondes 
had (1) a completely exhausted Bourdon tube barometer, which was 
found to show less fatigue effect than the aneroid barometer ; (2) Teisserenc 
de Bort’s bimetallic thermometer and Hergesell’s German-silver tube ther- 
mometer ; (3) a hair hygrometer.. The working parts of the instrument 
were enclosed in an aspiration tube. Similar mstruments were designed 
for use with kites. 
The principal self-recording instruments which have at various times 
been used have been designed by Richards Fréres, C. F. Marvin, 
Fergusson, L. Teisserenc de Bort, R. Assmann, H. Hergesell, and W. H. 
Dines.* Richards, Marvin, Hergesell, and Dines designed instruments for 
use with kites; and Richards, Teisserenc de Bort, Assmann, Hergesell, 
and Dines ballons-sondes instruments. 
Kite Meteorographs.—The Richard kite meteorograph is a_baro- 
thermo-hygro-anemograph. The barometer is a double aneroid, and the 
thermometer a Bourdon tube filled with alcohol. The hygrometer con- 
sists of a bundle of hairs, and the anemometer is of the Robinson cup 
type, operating through cogwheels and pulleys. The cups are mounted 
on a vertical spindle, projecting below the instrument, and thus 
work in an inverted position. The records are traced side by side 
on a smoked sheet fixed round a clockwork rotating drum, by styles 
connected through systems of levers to the various parts of the instru- 
ment. 
The whole instrument, with the exception of the anemometer cups, 
is enclosed in a protecting case. Ventilation is obtained by an opening 
in the front of the case and perforations in the back. The instrument 
is kept head to wind by means of a vane. Its total weight is 1,820 gm. 
The Marvin kite instrument is similar to the Richard instrument. 
The barometer is a large double aneroid with steel boxes. The ther- 
1 Compt. rend., 125, 1897, pp. 587-589. 2 Nature, vol, lxvii. p. 137. 
% A detailed account of Dines’ instruments and methods is contained in The 
Free Atmosphere in the Region of the British Isles, M.O., No. 202. 
