202 PROFESSOR THOMAS GRAHAM's SCIENTIFIC WORK. 



mercury within the tube uutil it stood at a height of 100 millimeters 

 only — that is, until the external pressure exceeded the internal pressure 

 by 100 millimeters only. Matters being in this state, the experiment 

 consisted in observing the number of seconds required for the admission 

 through the graphite septum, into the graduated tube, of a given 

 volume of gas — the mercury in the tube being kept throughout at the 

 constant height of 100 millimeters, by a gradual lifting up of the tube, 

 effected by a mechanical arrangement originally devised and employed 

 by Professor Bunsen. The long tubes were filled with mercury in a dif- 

 ferent manner; but the conduct of the experiments made with them 

 differed only from that of the experiments made with the shorttubes,in 

 that the level of mercury in the long tub(?s was maintained throughout 

 at or near to the barometric height, so that the external gas diffused 

 into the tube under full atmospheric pressure. Experimenting in this 

 way, the relative times of permeation of equal volumes of different 

 gases were found to be almost identical with the square roots of the 

 specific gravities of the respective gases, as shown in the following 

 table : 



Times of equal 

 difiusion. 



Square roots of 

 specific gravities. 



Oxygen 



Air 



Carbonic gas . 

 Hydrogen 



These results are of great value from the simplicity and constancy of 

 the conditions under which they were obtained, and from their close ac- 

 cordance with the induced law. By allowing the diffusion to take place 

 into a complete or partial vacuum, instead of into an atmosphere of 

 other gas, the results were not complicated with those of interditTusion; 

 and by employing a thin plate of highly compressed graphite, instead 

 of a comparatively thick plug of more porous stucco, the results were 

 not complicated with those of transpiration, as happened in some other- 

 wise admirable experiments of Professor Bunsen, which led that dis- 

 tinguished investigator to question at one time the accuracy of Mr. Gra- 

 ham's law. 



The absence of any transpiration of gas through the graphite wafer 

 was made evident by the want of any approximation, in the rates of 

 passage, to the characteristic rates of transpiration ; and was conse- 

 quent on the impermeability of the exceedingly minute pores of the 

 graphite to any enforced bodily transmission of gas through them. It 

 may be as well to state this conclusion in Mr. Graham's own words : 



"The movement of gases through the graphite plate appears to be 

 solely due to their own proper molecular motion, quite unaided by trans- 

 piration. It seems to be the simplest possiblei exhibition of the mole- 



