826 Intelligence and Miscellaneotis Articles. 



afterwards took it up, using designedly-cracked glasses, and demon- 

 strated that if, on the one hand, the hydrogen escapes outwards, the 

 air, on the other, travels in an opposite direction and penetrates into 

 the glass. By placing over the apparatus an uncracked bell-glass, so as 

 to circumscribe the exterior atmosphere, Magnus also ascertained that 

 the pressure increased in the air whilst it diminished in the hydrogen. 

 There were therefore here two gases in contact tlirough a capillary 

 passage, both of which traversed this communication in opposite 

 directions so as to unite ; and the hydrogen filtering more rapidly 

 than the air, a difference of pressure was soon proved. This was, 

 for gases, the phsenomenon of endosmose which Dutrochet had de- 

 tected in liquids. 



Since this period other facts of the same kind have been added to 

 those just referred to. Marianini having placed a soap-bubble in- 

 flated with air in carbonic acid gas, saw it increase in size at first, 

 and then burst. Other experimenters have produced the endosmose 

 of gases through porous vessels or thin partitions, and the general 

 fact has been placed beyond all dispute, although no convenient pro- 

 cess for exhibiting it has been discovered, and nothing has led to a 

 suspicion of the remarkable energy with which it is produced. I 

 have to lay before the Academy a certain means of proving this en- 

 dosmose and studying it in all its details. 



I take a porous vessel of unglazed porcelain intended for the Bun- 

 sen's batteries, wash it with alcohol, and after leaving it to dry for 

 several days, lay on its surface a coat of collodion dissolved in aether, 

 or gutta-percha in sulphuret of carbon. This coat must be very 

 thin, and envelope the entire surface of the vessel with a very con- 

 tinuous and equal varnish. I close it with a cemented obturator, 

 which gives passage to two tubes, of which one bears a stopcock, 

 whilst the other is open at both ends, and about 3 metres in length. 

 I fix the whole vertically on a support, when the apparatus exhibits 

 the porous vessel at its upper part, and the two tubes descend verti- 

 cally from it; the longer of these dips into a trough filled with 

 water, and the other, which bears the stopcock, serves for the intro- 

 duction of the gases to be studied. I now pass in a current of 

 hydrogen, which circulates in the porous vessel, returns by the long 

 tube, and escapes through the water surrounding its base. This 

 current of gas must be very abundant, and be maintained for several 

 minutes ; the stopcock is then closed. 



At this moment the pressure diminishes in the porous vessel, the 

 level of the water rises with remarkable rapidity, and in twenty 

 seconds it attains a maximum height which varies from 2 to 2*50 

 metres. The liquid soon sinks gradually, and in a few minutes the 

 internal pressure has again become equal to that of the atmosi)here. 



In its intensity, and almost in its rapidity, this experiment is equal 

 to that with cracked glasses ; it sliows that two unequal gaseous 

 currents are established, and that the difference of pressure which is 

 their consequence may exceed a fifth, and attain nearly a fourth of 

 an atmosphere; and as, at the moment when the maximum is ob- 

 served, the gases are already mixed, the effect perceived is inferior 

 to that which would be produced if these gases had remained pure. 



