68 PHYSIOLOGY OF NUTRITION. 



receive each four peas germinated in sawdust, others four oat seed- 

 lings each. The cylinders are placed at a window looking to the 

 south, and care is taken to replace daily the water lost by evapo- 

 ration. The pea plants are supported by sticks. The peas having 

 survived the starvation stage already referred to, thrive well, 

 while the oats keep sickly. In cylinders, however, in which the 

 sand receives an admixture of - 250 gr. KN0 3 over and above 

 the substances above specified, the oats nourish. The reasons for 

 these differences in behaviour are found in what has gone before. 



1 See Hellriegel, Beilageheft zu der Zeitschrift d. Vereins f. Riibenzucker- 

 industrie d. Deutschen Reiches, November, 1888. 



2 Such investigations are to be made according to the method of Boussingault 

 (see Agronomic, i. , p. 69). See also 19, above. 



3 Kegarding the nitrogenous food of the PapilionaceaB, Graminese and other 

 plants, I fully agree with the views of Hellriegel and Prazmowski (see Versuchs- 

 stationen, Bd. 37 and 38). Those of Frank, on the other hand (see Landwirth- 

 schaftl. Jahrbiicher, and Botan. Zeitung, 1893), I cannot share ; at all events 

 the experiments of Frank are not thoroughly convincing. 



21. The Detection of Ammonia and Nitric Acid in Water and 

 in Plants. Nitromonas. 



In order to satisfy ourselves that there occur in nature in- 

 organic nitrogenous compounds (ammonia, Nitric acid) which 

 can be utilised by plants in nutrition, it is well to test water 

 (river, pond, or spring water) for both these substances. For 

 ammonia we employ Nessler's solution. Two gr. of Potassium 

 iodide are dissolved in 5 c.c. of water, the solution is heated, 

 and to it we add rather more Mercuric iodide than it can 

 dissolve. The solution after cooling is diluted with 20 c.c. of 

 water, filtered and treated with potash solution (30 c.c. of potash 

 solution, consisting 1 part of caustic potash to 2 parts of water, 

 is added to 20 c.c. of the filtrate). We now fill two test-tubes 

 with the water to be examined, treat with soda lye, filter if 

 necessary, and add to the water in one tube about 30 drops of 

 Nessler's reagent. A comparison of the tint of the fluids in the 

 two tubes shows whether ammonia is pi^esent or absent, because 

 in presence of ammonia the fluid becomes reddish in colour. To 

 detect Nitric acid in water, a drop of the water is placed in a 

 white porcelain dish, and treated with 2 drops of a solution of 

 brucine (prepared by shaking up brucine in water). A few drops 



