MEMOIRS OF THE NATIONAL ACADEMY OF SCIENCES. 425 



Tlio alkalint' iiermauganate solution is tliorougiily boiled when fast pieiiared and made np to 

 the proper volmue. 



About 40 (•('. are used for eac^h 500 cc. deteruiination. 



This (piantity is steamed in a small liask ou a hot, iron plate during the collection of the free 

 ammonia distillates and is turned into the apparatus from this. Eight of these flasks are used, all 

 heated a1 once ou the jdate. In the case of the 100 cc. determinations about 50 cc. of the perman- 

 ganate solution is divided into two nearly equal portions between (wo of these small flasks and 

 steamed in the same manner. The portion contained in each of the flasks is afterwards divided, 

 as equally as may be, between two of the small retorts by means of a pipette. The distillates in 

 the 100 cc. mediod are collected in small test tubes in portions of about 10 cc, three portions for 

 the " free ammonia '" and hve for the " albuminoid ammonia." These test tubes are about 11 mm. 

 in internal diameter and about 14 cm. long. A volume of 10 cc. fills one of them to a height of 

 about 10 cm. 



The standards used in this method are made up in the same tubes with b(jiled distilled water, 

 to about the same volume as the distillates, and contain, respectively, 0.05, 0.1, 0.15, 0.2, 0.3, 0.4, 

 0.5, 0,0, 0.7, 0.8, 0.9, 1, 1.1, 1.2, 1.3, and 1.4 cc. of the amnu)nium chloride solution used in the 

 other method. In nesvslerizing about 0.5 cc. of the reagent is used for each tube. No difficulty is 

 experienced in reading to 0.0005 mg. of ammonm. Many of the details here adopted have been 

 derived from the papers of Drown, above mentioned, and of Erenneman. (Jour. Amer. Chem. 

 Soc, VIII, No. 9, p. 221.) 



The volumes of 50 and 10 cc. mentioned are not measured accurately, but are only approx- 

 imate. 



The nitrogen as nitrates has been determined by the method of Grandval and Lajoux (Comptes 

 Ecndus, July 6, 1885) as modified by Drown and Hazen (Eep. Mass. Board of Health, 1890, pt. 2, 

 p. 711). Five cubic centimeters of the water is placed in a porcelain dish of about 35 cc. capacity, 

 two or three droi)s of a 1 per cent solution of sodium carbonate added and evaporated to dryness 

 on the top of a water bath. The steam should not be allowed to come in contact with the dish. 

 The residue of the evaporation is treated with about 0.5 cc. of phenol suli)huric acid, made by 

 digesting about 23 grams of pure crystalized phenol in 200 cc. of pure sulphuric acid for some 

 hours. By ai)i)ropriate manipulation of the dish the aiad is worked well over the bt)ttom and 

 sides. After some time a few centimeters of distilled water are added and then a solution of 

 caustic potash until the yellow color is well brought out. The strength of the caustic potash 

 solution should be 10 per cent or more. It is imi)ortant that too great an excess of this reagent 

 be not added, for if this occurs crystals of potassium sulphate are thrown down, a result which is 

 not desirable. A set of standards is made up for each analysis from 1, 2, 3, 4, 5, and C cc. of a 

 solution of potassium nitrate, of which 1 cc. contains 0.001 mg. of nitrogen as nitrate. The first 

 and last standards have often been omitted. These quantities are placed in small porcelain 

 dishes, of the capacity above mentioned, and treated in the same manner as that described for 

 the sample. In making up the standard solution 0.7215 gram of potassium nitrate is dissolved 

 in 1,000 cc. of water, and this solution diluted 1:100. 



The samples and standards are evaporated and treated together as above described, and then 

 the contents of the dishes are transferred to the long tubes used iu the ammonia determinations. 

 More distilled water is added until they are all filled to about 2 or 3 cm. from the top, and iu 

 these the reading is made. The determination is thought to be ino.st accurate when the amount 

 of nitrogen as nitrates iu the (juantity of water taken falls between the limits of the standard 

 here employed. The tests have been usually made in duplicate, and the evaporation of eight of 

 them — two for each of the four samples of water — and a set of standards lias been simultaneous. 



The regular determination of the nitrogen as nitrites was stojtped in January. The method 

 was essentially that described by Drown (Rep. Mass. Board of Health, 1890, pt. 1, p. 527), and 

 called by him "Warrington's modification of the Griess method." 



The water was decolorized before the test by treatment with a few drops of solutions of alum 

 and of sodium carbonate, of about 10 and 33 per cent, respectively, and then filtering through 

 Alter paper, which had been washed to remove nitrous acid. 



