MEMOIRS OF THE NATIONAL ACADEMY OF SCIENCES. 7 



The most convenient tube for use in this jar is <> inches long by three- eighths of an inch in 

 diameter. 



A word may be said about hydrogen generators. The apparatus of Kipp lias a wide use, but 

 has not proved altogether satisfactory where a considerable quantity of the gas is desired. The 

 apparatus of Sternberg is an excellent one, and simple in construction (Fig. 7). (Manual of 

 Bacteriology, p. S3.) I have found much satisfaction in the* use 

 of an apparatus constructed as follows (Fig. 8): Two bottles hold, 

 ing from 3 to 4 liters are fitted with perforated rubber stoppers, 

 that of bottle A carrying two tubes, one reaching to the bottom 

 of the bottle, while the other is short, bent at right angles, and 

 fitted with a stopcock. The stopper of bottle 1! also carries a 

 tube reaching to the bottom, and into the second hole a glass plug 

 is placed, for which a safety tube may be substituted. By the use 

 of the plug the gas can be stored in greater quantity under some 

 pressure. Broken glass is put into bottle A for a depth of about ^ 

 an inch, and on this the zinc is placed. The long tubes are con- 

 nected with stout rubber tubing, on which it is well to have a screw 

 pinchcock. Bottle B is raised some 4 inches above A, and the 

 dilute acid placed in it. To start the apparatus, slight suction is 

 applied to the outlet tube of bottle A. As soon as the acid starts 

 to flow it will go by the action of the siphon into bottle A. When 

 enough has gone over, the pinchcock is closed to prevent the suc- 

 tion of air. On closing the stopcock of the outlet tube the acid is 

 forced back into the upper bottle, and the zinc is left dry. By using 

 a stopper in the upper bottle instead of a safety tube, there is 

 always enough pressure to force the acid over into A, making the 

 apparatus practically automatic. As it is almost always used in connection with an air pump, 

 however, this is not a matter of importance. The stopcock of the outlet should never be turned 

 off when the tube connecting the two bottles is closed, otherwise an explosion is apt to result. 

 With this apparatus one has always at hand from 3 to 4 liters of hydrogen under pressure, antl it 

 is a simple matter to till any vessel, from a test tube to a bell jar. Iu every case the air is first 

 exhausted, and then the hydrogen let in, the process being repeated three or more times. 1 



Fig. 7. 



For the purpose of exhausting the air, the most convenient pump is the simple one constructed 

 on the principle of Sprengel's mercurial pump, in which a stream of water does the work. It is 

 attached to a spigot, and the suction tube is connected with a Woulf bottle to prevent any 

 back flow of water. 



The entire apparatus, including the anaerobic jar, is shown iu Fig. 8. 



1 This apparatus was suggested to me by Dr. S. S. Kneass, a former student iu Professor Roux's laboratory at 

 the Pasteur Institute, where he saw a similar one in use. 



