'THE VITAL PHENOMENA OF BACTERIA. 109 



ammonia before it can enter into the proteid 

 synthesis. Frankland and Low have in this 

 connection corrected my own earlier concep- 

 tion. I had found no nitrous acid as an inter- 

 mediate stage and had therefore supposed a 

 direct oxidation of the ammonia to occur : 



I. NH 3 -+- 20 2 = HN0 3 + H 2 

 ammonia, oxygen, nitric acid. 



II. C0 2 + H 2 = H 2 C0 3 = CH 2 + O 2 

 carbon dioxide, water, carbonic acid, formaldehyde, oxygen. 



while Low supposes : 



I. 2NH 3 + 2O 2 = 2HNO 2 + 2H 2 

 ammonia, oxygen, nitrous acid, hydrogen 



II. C0 2 + 2H 2 = CH 2 4- H 2 

 carbon dioxide, hydrogen, formaldehyde, water. 



The further synthesis of carbohydrates and 

 proteids is carried out in the way previous- 

 ly stated. Perhaps the possibility of both 

 methods of oxidation should be taken into con- 

 sideration, for I at least entirely overlooked 

 the nitrous acid, while Winogradsky now 

 supposes that two kinds of bacteria exist, 

 the one of which oxidizes ammonia only as far 

 as nitrous acid, and the other oxidizes nitrous 

 acid to nitric acid. According to Winogradsky, 

 the oxidation of 35.4 milligrams of nitrogen is 

 necessary in order to bring i milligram of 

 carbon into organic combination. The matter 

 has a practical interest since, by the use of 

 material which contains the nitrifying bac- 

 teria as a fertilizer, it is possible to increase 



