On the Dialytic Separation of Gases by Colloid Septa. 503 



Four of the above coefficients differ considerably from their 

 values found by observation: it will be found that they receive 

 corrections when the approximation is extended to terms of the 

 fourth order. 



The facility of the method of computation above given can be 

 appreciated only by those who have actually gone through the 

 investigation according to the usual methods. I purpose, in 

 a future paper, to carry the approximation as far as terms of the 

 third and fourth order, and to remark on one or two points 

 connected with the method which require elucidation. 



LXIX. On the Absorption and Dialytic Separation of Gases by 

 Colloid Septa. By Thomas Graham, F.R.S., Master of the 

 Mint. 



[Continued from p. 420.] 



Part II. — Action of Metallic Septa at a red heat. 



Platinum. 



rHE surprising passage of gases through the homogeneous 

 substance of a plate of fused platinum or of iron at a red 

 heat, lately discovered by MM. H. St. -Claire Deville and Troost, 

 may possibly prove to be analogous in its mode of occurrence to the 

 passage of gases through the rubber septum. At the same time 

 it must be admitted that such an hypothesis as that of liquefac- 

 tion can only be applied in a general and somewhat vague man- 

 ner to bodies so elastic and volatile at an elevated temperature as 

 the gases generally must be, and hydrogen in particular. Still 

 some degree of absorbing and liquefying power can scarcely be 

 denied to a soft or liquid substance, in whatever circumstances 

 it may be found, with such a patent fact before us as the reten- 

 tion by fused silver of 18 or 20 volumes of oxygen at a red heat. 

 It may safely be assumed that the tendency of gases to liquefac- 

 tion, however much abated by temperature, is too essential a 

 property of matter to be ever entirely obliterated. 



A little consideration also shows that the absorption of gas by 

 a liquid or by a colloid substance is not a purely physical effect. 

 The absorption appears to require sonfe relation in composition 

 — as where both the gas and the liquid are hydrocarbons, and 

 the affinity or attraction of solution can come into play. May a 

 similar analogy be looked for of hydrogen to liquid or colloid 

 bodies of the metallic class ? 



With reference to the mechanical pores of a solid mass, liquids 

 are probably more penetrating than gases. The former show 

 often a power of adhesion to solids, while gases appear to be es- 



