The instrument was suspended within the receiver of an airpump, over a cup containing water at the temperature of 62°, the mercurial barometer standing at 30 inches. The air in the receiver being exhausted to a degree of rarefaction corresponding to twenty-nine inches of the barometer, the lower end of the instrument was immersed in the cup of water; and air being admitted into the receiver, the exhaustion was repeated until the barometer gauge indicated a pressure equal to twenty-eight inches, when a corresponding mark was made on the tube, the air being in like manner admitted after its re-immersion. By the repetition of this process, the graduation of the stem was carried on as far as was

necessary.

With several tubes thus graduated, an observer in a hilly country may ascertain the density of the atmosphere on the summits of different elevations, by sending an assistant to each, with one of these tubes, and a tin case containing water. They are taken up with the stems open; and the air within each partaking of the density of that at the station, the mouth of the tube is put into the water, and left in it as the assistant descends. The water will rise in the stem as the density of the atmosphere increases, and will indicate by its height the degree of rarefaction of the air at the upper station-a correction being made for the variation of the barometer from the standard height, and also for that of the temperature of the atmosphere.

This substitute for the expensive and delicate mercurial mountain barometer would, from its portability and simplicity, be particularly useful in determining comparative altitudes in a mountainous country, but of course the same accuracy cannot be expected from it.Another method of obtaining approximate differences of altitude is by a comparison of the temperatures of boiling water (which vary with the pressure of the atmosphere), upon which a paper was some years since published by Colonel Sykes, who practised it extensively in India *.

As the necessary apparatus is exceedingly simple, and the in

* I ascertained lately the approximate altitudes above the sea of a number of places în Australia by this method; many of these were afterwards tested by the triangulation, and the results proved even more satisfactory than I had anticipated.

strument not so liable to injury as the mercurial barometer, and much more portable and easily replaced, I have taken from this paper, which will be found in the 8th number of the "Geographical Journal," the tables computed by Mr. Prinsep, to facilitate the computation of altitudes, and also the examples given by Colonel Sykes, which render their application evident without further explanation.

The results deduced from the use of these tables appear always rather less than those obtained from careful barometrical observations, and also less than those calculated from the different formulæ, which have been arranged for the determination of altitudes by this method, but which do not all agree. The results of a number of careful observations made with the thermometer, compared with those obtained at the same time with the barometer; or which have been ascertained by levelling, or trigonometrically, will afford the means of making any necessary corrections in the tables; which, however, giving so close an approximation, deserve to be more generally known and made use of.

The accompanying sketch and explanation, taken from Col. Sykes's pamphlet, show the whole apparatus required : A. A common tin pot, 9 inches high by 2 in diameter.

D

B. A sliding tube of tin, moving up and down in the pot: the head of the tube is closed, but has a slit in it, C, to admit of the thermometer passing through a collar of cork, which shuts up the slit where the thermometer is placed.

D. Thermometer, with as much of the scale. left out as may be desirable.

E. Holes for the escape of steam.

The pot is filled four or five inches with pure water; the thermometer fitted into the aperture in the lid of the sliding tube, by means of a collar of cork; and the tin sliding tube pushed up or down to admit of the bulb of the thermometer being about two inches from the bottom of the pot.

Before using a thermometer for this purpose, it is necessary to

LE B

ascertain if the boiling point is correctly marked for the level of the sea by a number of careful observations, and the difference, if any, must be noted as an index error. It is always desirable to have two or more thermometers which have been thus tested; and in all observations the temperature of the air at the time should be noted.

Boiling Point of Water.

TO FIND THE BAROMETRIC PRESSURE AND ELEVATION CORRESPONDING TO ANY OBSERVED TEMPERATURE OF BOILING WATER BETWEEN 214° AND 180°.

O

214

213

212

211

210

209

208

207

206

205

204

203

202

201

200

199

198

197

196

195

194

193

192

191

190

189

188

187

186

185

184

183

182

181

180

TABLE I.

92.6

93.0

93.4

93.8

94.2

94.8

95.3

95.9

96.4

96.9

97.4

97.9

The Fourth Column gives the Heights in Feet.

Temperature of the Air.

TABLE II.

TABLE OF MULTIPLIERS TO CORRECT THE APPROXIMATE HEIGHT FOR THE TEMPERATURE

OF THE AIR.

Multiplier.

Temperature of the air above.
Ditto ditto

below..

1.042

1.044

1.046

1.048

1.050

1.052

1.054

1.056

1.058

1.060

1.062

1.064

1.066

1.069

1.071

1.073

1.075

1.077

1.079

1.081

Temperature
of the Air.

When the water (with the thermometer immersed) has been boiled at the foot and at the summit of a mountain, nothing more is necessary than to deduct the number in the column of feet opposite the boiling point below, from that opposite the boiling point above: this gives an approximate height, to be multiplied by the number opposite the mean temperature of the air in Table II., for the correct altitude.

Boiling point at summit of Hill Fort of
Púrundhur, near Púna.

Boiling point at Hay Cottage, Púna

90

91

Multiplier.

1.083

1.085

1.087

1.089

1.091

1.094

1.096

1.098

1.100

1.102

1.104

1.106

1.108

1.110

1.112

1.114

1.116

1.118

1.121

1.123

Approximate height 2337

75°

83

Feet.

204.2

4027

208.7 1690

Mean 79 Multiplier 1.098

Correct altitude 2566 feet.

I

When the boiling point at the upper station alone is observed, and for the lower the level of the sea, or the register of a distinct barometer is taken; then the barometric reading had better be converted into feet, by the usual method of subtracting its logarithm from 1·47712 (log. of 30 inches) and multiplying by 6, as the differences in the column of “ barometer" vary more rapidly than those in the "feet" column.

Assuming 30-00 inches as the average height of the barometer at the level of the sea (which is however too much), the altitude of the upper station is at once obtained by inspection in Table I., correcting for temperature of the stratum of air traversed, by Table II.

In moderate elevations, the difference of one degree in the temperature at which water boils, indicates a change of level of about 500 feet, nearly equivalent to what would be shown by a difference of 0.6 of an inch in a mercurial barometer.