74 REPORTS ON THE STATE OF SCIENCE. 
of height: a result contrary to that of Gay-Lussac, but in agreement 
with that of Robertson. 
In 1869 Glaisher made further observations to examine more closely 
the variation of temperature and humidity up to 1,000 feet. These later 
observations were the first obtained by means of a captive balloon. They 
indicated a decided diurnal range of temperature.! The work was not 
pursued further in England, but interest was stimulated in France, and 
many ascents were made by MM. Flammarion, de Fonvielle, and Tissan- 
dier. Jn 1875 two ascents were made by Tissandier, Crocé-Spinelli, 
and Sivel, one of long duration (24 hours) and the other to a great height 
(9,000 metres). The apparatus carried included a pump to draw air through 
tubes filled with potash for estimating the amount of carbon dioxide, a 
spectroscope for examining the water-vapour line in the solar spectrum, 
two aneroid barometers, one giving the pressure from 0 to 4,000 m. and 
the other from 4,000 m. to 10,000 m., two barometric tubes for registering 
the lowest pressure, and thermometers.” This ascent resulted in the 
death by suffocation of Crocé-Spinelliand Sivel, and in consequence only 
one ascent was made in France between 1875 and 1878. The successful 
construction of the large Giffard captive balloon in Paris in 1878 gave a 
great impetus to aeronautical work in France.* In 1879 the Paris Academy 
Inaugurated its first series of ascents, and Tissandier made observations 
to verify the barometer height formule by means of photographs from 
the car, a method originally proposed by Le Verrier in 1874.4 In the 
same year (1879) the International Congress of Meteorologists at Rome 
passed several resolutions relating to the importance of balloon observa- 
tions in meteorological investigation.” In July 1881 MM. W. de Fonvielle 
and Lippmann made an ascent after midnight, carrying only barometer 
and thermometer,® and about the same time the investigation was 
resumed in England by the Meteorological Office.’ 
The number of balloon ascents accomplished and the observations 
made had now become very numerous and widely distributed, but it was 
seen that the results were strangely discordant. No organised balloon 
ascents had taken place in Germany, though isolated ascents had been 
made since about 1880; but German meteorologists attributed the 
discrepancies to faulty instruments and methods of observation. Com- 
parisons by A. L. Rotch in ascents from Paris and Berlin showed that 
Richard’s self-recording thermometer registered 8° C. higher than a sling 
thermometer, and the latter 2° higher than a new aspirated thermometer 
designed by Assmann in 1887.8 These tests showed the necessity for the 
use of accurately specified methods of observation and thoroughly tested 
instruments in all countries. The doubt thrown on all previous obser- 
vations caused the Prussian Meteorological Institute to inaugurate a 
series of experiments to repeat the work of Glaisher. Forty-seven ascents 
were made between June 1888 and February 1895. In these ascents 
the instruments were placed in a well-ventilated enclosure and their indica- 
tions were compared with those of instruments exposed as in earlier 
ascents. In four cases self-recording instruments were used. On December 4, 
1894, Dr. A. Berson rose to a height of 9,600 m., the highest level at 
*<On the Changes of Temperature and Humidity of the Air up to 1,000 feet,’ 
J. Glaisher, Brit. Assoc. Reports, 1869. 
* Compt. rend., 1875, pp. 803, 866, 976, 1060. 
* Nature, vol. xviii. p. 291. * Tbid., vol. xviii. p. 160. 
> Tbid., vol. xx. p. 58. 6 Thid., vol. xxiv. 225. 
* Tbid., vol. xxv. p. 158, 8 Tbid., vol. xlv. p. 168; vol. xliv. p. 512. 
