496 Prof. Tyndall on the Action of Free Molecules on 



action upon radiant heat. Since 1864 this problem has been 

 often in my mind. The wide silvered tube, I am happy to 

 say, has rendered the solution of the problem possible. 



It is only highly volatile liquids that lend themselves to 

 this experiment, because from them alone can vapours be 

 be derived of sufficient density to produce liquid layers of 

 practicable thickness. On the 22nd of last October the experi- 

 ment was first attempted. The source of heat was the lime- 

 light, the rays of which were received by a concave mirror 

 silvered in front, and sent in a nearly parallel beam through 

 the experimental tube. At the end nearest the source the tube 

 was provided with a diaphragm having a circular orifice 

 1 inch in diameter. At the other end was a diaphragm with an 

 orifice \ an inch in diameter. Beyond this was placed the 

 thermopile, furnished, not with its reflecting cone, but with 

 a tube of brass (shown in fig. '6) 2 inches long and blackened 

 within. In this arrangement, the heat which reached the 

 pile did not even approach the interior cylindrical surface. 

 The total heat employed produced a deflection of 60 galvano- 

 metric degrees, which, when the tube was exhausted by a 

 powerful Bianchi's air-pump, was accurately neutralized by a 

 compensating cube. Liquid sulphuric ether was then placed 

 in a large flask provided with a sound stopcock, the object 

 being to expose a considerable evaporating surface. The flask 

 was plunged in water, with the view of keeping the liquid 

 and its vapour at an approximately constant temperature. 

 The air being carefully removed from the flask, it was attached 

 to the experimental tube, and a quantity of vapour was allowed 

 to enter sufficient to render a column 38 inches long equivalent 

 to a liquid layer 1 millimetre in thickness. Two concurrent 

 experiments made the deflection produced by the vapour 

 41°. 

 Without altering the quality of the heat, the absorption 

 exercised by a liquid layer of sulphuric ether 1 millimetre 

 thick was next determined. The rock-salt cell with which the 

 experiment was made is described in detail in the Bakerian 

 Lecture for 1864 (Phil. Trans, vol. 154. p. 328.) The an- 

 nexed figure (fig. 5) will give a sufficiently clear notion of its 

 construction and disposition. Between two stout plates of 

 brass, c e and its fellow, two rock-salt plates of extreme purity 

 are firmly clasped by suitable screws, due care being taken to 

 protect the plates from a crushing pressure. The two brass plates 

 referred to are perforated by circular apertures, as shown in 

 the figure. The two plates of salt are not allowed to come 

 into contact, but are separated from each other by a carefully 





