510 Prof. Tyndall on the Action of Free Molecules on 



reduce the diameter to 2*5 inches. Against these flanges, 

 transparent plates of rock-salt are fixed air-tight. The tight- 

 ness of the tube was secured, sometimes by india-rubber 

 washers properly greased, and sometimes by cement*. A 

 stop-cock a near one end of T T' was connected with a baro- 

 meter-tube and an air-pump. A T piece at the other end 

 was connected on the one side with a purifying-apparatus (not 

 shown), consisting of two U tubes, one containing fragments 

 of Carrara marble wetted with caustic potash, the other con- 

 taining fragments of glass wetted with sulphuric acid. Before 

 entering these U tubes the air was freed from suspended mat- 

 ter by a plug of cotton -wool. On the other side, the T piece 

 was connected with a quill tube of glass bent into the shape 

 of a U, in the two legs of which a coloured liquid stood at 

 the same level. The liquid column when standing at the same 

 level in both arms of the U was 350 millim. high in each, 

 while the free leg of the U rose to a height of about 500 mil- 

 lim. above the surface of the liquid (shortened in the figure). 

 The source of heat was the lime-cylinder L, rendered incan- 

 descent by a flame of coal-gas and oxygen. The rays from 

 the lime-cylinder were received by a concave mirror R silvered 

 in front, and sent by it in a convergent beam through the 

 manometer-tube. The focus of the beam was within the tube 

 and near its most distant end. The gas and oxygen were 

 supplied from gas-holders specially constructed for these and 

 similar experiments — long experience of the futility of gas from 

 the public mains, or compressed in iron bottles, having shown 

 independent gas-holders which could be kept at an unalterable 

 pressure to be essential. 



The experiments were conducted thus : — A test-tube t, 

 plunged in water held by the glass g, contained the liquid 

 whose vapour was to be examined. Through a cork which 

 stopped the test-tube passed a narrow tube of glass, ending in 

 a small orifice near the bottom of the test-tube, and at a con- 

 siderable depth below the surface of the liquid. To augment 

 this depth, and to economise the liquid, the lower part of the 

 test-tube was drawn out to half the diameter of its upper part. 

 A second narrow tube passed also air-tight through the cork, 

 and ended immediately beneath it. Both tubes were bent at 

 a right angle above the cork. The manometric tube being 

 exhausted, air freed from its carbonic acid, its moisture, and 



very day, moreover, on which I made my communication to the Society 

 of Telegraphic Engineers, viz. the 8th of December, 1880, he forwarded 

 to a scientific journal the announcement of bis having obtained sounds 

 from coal-gas and ammonia (see Wiedemann's Annalen, Jan. 1881). His 

 subsequent experiments with aqueous vapour &c. agree with mine. 



* The absorption of the vapours by india-rubber (which was in some 

 cases extraordinary) caused the washers to be abandoned. 



