54 



BACTERIA IN RELATION TO PLANT DISEASES. 



beef-boiiillon and steamed ever}^ twenty-four hours for seven or eight da}'s, a large 

 bubble being tilted out each time and appearing just as regularly during the next 

 steaming. Naturally, no strict anaerobe would grow in such a tube and every 

 aerobe would appear to be a facultative anaerobe. The neck of the fermentation- 

 tube should be as narrow as consistent with filling and cleaning. All wide-necked 

 tubes should be discarded. The behavior of the closed end wath reference to the 



absorption of air may be tested by adding litmus- 

 water and 5 per cent grape-sugar to the bouillon. 

 On steaming, the litmus is reduced. If there 

 is no air in the closed end the litmus remains 

 reduced, while in the open end exposed to the air 

 it soon oxidizes back to its original color. 

 Other things to be observed are ; 

 (j) Grozvth in hydrogen. 

 {4) Growth in carbon dioxide. 

 (5) Growth in vacuo., various degrees of ex- 

 haustion. 



{6) Grozvth in vacuo ^ remnant of oxygen ab- 

 sorbed by the mixture of caustic potash and pyro- 

 gallol (same as pyrogallic acid). 



(7) Groivth in nitrogen (air with the oxj-gen 

 absorbed, normal air-pressure). 



The hydrogen and carbon dioxide, which are 

 required in considerable quantities, may be gen- 

 erated in Kipp gas-generators. There is a choice 

 in generators. The writer has not found any kind 

 which is entirely satisfactor}-. The one which has 

 given the least trouble is shown in fig. 50. The 

 objection to this generator is the large volume of 

 dead acid which soon accumulates at the bottom. 

 The accumulation of dead acid is entirely obvi- 

 ated in the de Koninck generator, but the writer 

 has onl)' recently obtained this apparatus and has 

 not yet had enough experience with it to speak 

 unqualifiedly. It furnishes a large amount of gas 

 and its generation may be stopped vei-}' quickly, 

 biit the acid chamber is inconveniently bulky (10 

 liters) and in case of breakage a destructive flood 

 would be poured out into the laboratory. To avoid this the apparatus should be 

 set into a deep enameled iron pan. The action of the apparatus depends on the fact 



Fig. 50.' 



*FiG. 50. — Kipp gas-gcncrator fc^r making carbon dioxide or hydrogen. When not in use the 

 pressure of the gas forces the acid off the marble or zinc (in the middle compartment) and stops 

 its evolution. Much reduced. 



