xxiv 



vitality ; and to these gradual colloidal changes may be referred the cha- 

 racteristic protraction of chemico-organic changes. These colloid bodies 

 are very easily penetrated by the soluble crystalloid bodies to which they 

 are opposed, and they form a medium also of separation or dialysis. 



This process of dialysis has a high value in the explanations it affords, 

 and promises to afford, of many physiological phenomena. But these 

 hitherto unknown and still obscure properties of molecules seem destined 

 to lead us still further on ; and by presenting to us a nearer view of the 

 fundamental phenomena, they give us an idea of the enormous magnitude 

 of the structure under which they seem to lie. Graham believed that the 

 rate of diffusion held a place in vital science not unlike the time of the 

 falling of heavy bodies in the physics of gravitation. 



In a paper " On the Molecular Mobility of Gases," June 18th, 1863, 

 he further compares several substances as to their facilities for diffusion, 

 and defines clearly the effusion-rate and the transpiration-rate as distinct. 

 A substance suiting the purpose of diffusion is graphite. 



He obtained by the graphite diffusiometer a separation of oxygen from 

 the air, making a mixture with 2 per cent, additional of that gas. This led 

 him to try another mode ; and by lengthening the surface and tube, he 

 obtained 3| per cent, more oxygen than in the atmosphere. This process 

 he calls atmolysis. Trying diffusion without an intervening septum, he 

 found that carbonic acid had proceeded half a metre length in seven minutes. 



The separation of gases was carried out much further, and described in 

 a paper read June 21st, 18G6. Here he begins the use of caoutchouc, 

 having been led to it by the experiments of Dr. Mitchell, of Philadelphia. 

 He found that air drawn through sheet rubber contained as much as 41*8 

 oxygen, the theoretical speed being 40*46, deduced from the passage of 

 the separate gas. He is desirous of showing that in this case the flow is 

 different from diffusion ; it is caused by an absorption of the gases, which 

 are taken into the caoutchouc in a liquid state, and are then given out on 

 the opposite side. 



This inquiry naturally led to an examination of the absorption by 

 metals. Deville and Troost had discovered that platinum and iron absorbed 

 gases when hot. Graham found hydrogen to pass through heated platinum 

 1*1 millim. thick at the rate of 489*2 cub. centimetres per minute on a 

 square metre. Oxygen scarcely passed, and other gases tried did not 

 pass. Wrought platinum took up 5*53 vols, of hydrogen, which, on 

 cooling, were shut up or occluded in the mass. Fused platinum took only 

 0*171 vol.; hammered platinum 2*28-3*79. Palladium, however, was 

 most remarkable, as it took up 643 vols, of hydrogen ; in a later paper 

 the quantity is stated to be 935 vols. These gases were pumped out from 

 the reheated, but could not be removed from the cold metal. Palladium 

 cold, however, was found to take up hydrogen when it was used as the 

 negative pole of a galvanic battery, and spongy palladium, which had ab- 

 sorbed hydrogen when heated, deoxidized some salts in the cold. 



