404 H. M. Vernon 



organisms, or are the ebanges of same other nature, such as a 

 reducing one? The frequent occurrence of Ditrifyìng organisms in 

 fresh water has been deraonstrated b}^ Uffelmann, Adametz, and 

 others, whilst Gayon and Dupetit, and also Heraeus, bave shown 

 the existence of denitrifying organisms*. In the present instance, it 

 was thought sufficient to determine only the nitrites present in the 

 water, as a fair eriterion of the nitriiication taking place can thereby 

 be obtained, and moreover their determination is so very much simpler 

 than that of nitrates. The method adopted was that of Preusse and 

 Tiemann2. This is a quantitative colorimetrie method depending 

 on the yellow colouration, due to the formation of triaraido-azobenzol, 

 produced on the addition of metaphenylene-diamine solution to the 

 acidified water. Standard solutions containing 1, 2, 3, 5 and 10 ce. 

 of a .001^ solution of N2O3 in 100 ce. of water, were freshly made 

 up each time, and by comparing various thicknesses of these solutions 

 with the water under examination, by meaus of the colorimeter 

 previously described, fairly accurate determinations were made. In 

 that the colour of the standard solutions is not arithmetically pro- 

 portionate to the amount of nitrite present, and in that it increases 

 gradually on keeping, the method can, however, yield only approximate 

 results. 



Determinations made at intervals of a week or two during the 

 months of September and October showed the normal Aquarium tank 

 water to contain .054, .076, .121 and .108 mgm. of N2O3 per litre, 

 or on au average, .090 parts per million of water. A specimen of 

 sea-water collected 5 km. from the shore contained .022 mgm. 

 Specimens of water kept in darkness for 35 days were fouud to 

 contain only .013 and .018 mgm.; those kept 46 days .027, .00, and 

 .009 mgm.; and a specimen kept 66 days .00 mgm. Thus it follows 

 that so far from the bacteria oxidising the ammonia present in the 

 water to nitrites, they remove a considerable proportion of those 

 already present. It is of course possible that the oxidation may 

 bave progressed as far as the nitrate stage, but if so, the above 

 results show that this process must bave been nearly corapleted after 

 35 days. That the water may contain nitrifying organisms is shown 

 by the analysis of a specimen of water Avhich had been heated to 

 100" C. and then kept 46 days in darkness. At the end of this time. 



' Untersuchuug des Wassers p. 497. 



■2 Ber. D. Chem. Ges. 11. Bd. 1878 p. 627. 



