424 MEMOIRS OF THE NATIONAL ACADEMY OF SCIENCES. 



tubes. Up to this date 500 cc. of water were taken from every determiuatiou, but beginning- 

 with tlie analyses of April 17 the use of the flasks was abandoned and small retorts of 200 

 to 500 cc. capacity substituted. In these 100 cc. of water were used for each determination, while 

 500 cc. continued to be the quantity taken for charging each of the larger retorts, the capacity of 

 which has varied from about a liter to a liter and a half. This arrangement has been continued up 

 to the close of the work. 



In making the duplicate determinations a flask and a retort, or the 500 cc. and the 100 cc. 

 methods, were used for each sample, so that comparisons might be made of the results obtained by 

 both types of apparatus or by both uu'thods. In the table of analyses the columns of ammonia 

 determinations marked "tin tube" and "glass tube," respectively, contain the results given by 

 the use of tin and glass condensing surfaces. 



The tin tubes were used on the supposition that better condensation would be obtained and 

 therefore higher results, but this does not aeem to be supported by the figures. In the same 

 columns, but farther down, the determinations by the 100 cc. method maybe compared witli tiiose 

 of the 500 cc. method. It will be seen that as a general thing the former method gives somewhat 

 higher results. 



The retorts and flasks are supported on iron rings and wire gauze. The apparatus is prepared 

 for each determination by charging with distilled water and distilling off a quantity, the later 

 portions of which are tested for anmionia, and if negative or if only a slight trace of color is 

 produced by the "nessler" the residue is removed and the proper quantity of the sample turned 

 in. For some time sodium carbonate was added to the water before beginning the distillation for 

 free ammonia, but this has long been discontinued. The presence of this salt seems to favor 

 explosive "bumping" and to retard ebullition. The distillation is usually made within twelve 

 hours after the collection and carried out in the afternoon or evening. About the same time a. 

 series of standards is made up and these, together with the distillates, are nesslerized the next 

 morning. 



The rate of distillation has varied, as a constant rate could not very well be obtained when 

 so many pieces of apparatus were working at once. 



In 500 cc. determinations glass tubes 31 cm. long, which a volume of 50 cc. fills to within 

 several centimeters of the top, are used for containing the standards and distillates, and in these, 

 in volumes of about 50 cc, the readings are made. The standards consists of eighteen, of these 

 tubes, in which are placed, respectively, 0.1, 0.2, 0.3, 0.4, 0.5, 0.7, 0.9, 1, 1.3, 1.5, 1.8, 2, 2.3, 2.4, 2.5, 

 2.8, 3 3.5, and 4 cc. of the standard ammouium chloride solution, of which 1 cc. contains 0.01 mg. 

 of ammonia and sufficient distilled water to fill them to within some centimeters of the top. 



The distilled water used for this purpose from January onward lias been boiled. For the 

 "free ammonia" 150 cc. are distilled. This is collected either in three portions of 50 cc. each or 

 in a first portion of 100 cc. and a second por(,ion of 50 cc. For the albuminoid ammonia 250 cc. 

 are collected, the first 100 cc. in one portion and the remainder in tiiree jiortions of 50 cc. each. 

 The 50 cc. distillates are received in the long tubes above described. All the distillates and 

 standards are nesslerized the next morning at one time, about 1 cc. of Nessler's reagent being 

 used for each. The determination of the amount of ammonia in the 100 cc. portions is effected 

 by transferring each into two of the long tubes after nesslerization and comparing these with the 

 standards, as in the case of the 50 cc. distillates. This mode of collecting some of the distillates 

 in 100 cc. portions was adopted not only because there were not enough of the long tubes, but 

 also because the first 50 cc. of both kinds contain a large part of the ammonia, and as this 

 quantity may be larger than that contained in the highest standard, it can be estimated with 

 greater convenience if distributed through two volumes of 50 cc. A specimen determination of 

 albuminoid ammonia is here given in detail to show the distribution of the ammonia through the^ 

 distillates. The collection and reading has been carried out as above described. 



( 0.0215 mg. N II3 



(1) 100cc.= (^,_„oi5ajg.NH3 



(2) .50 cc. = 0.0035 mg. N H3 



(3) 50 cc. = 0.0025 mg. N H3 



(4) 50 cc. = 0.0040 mg. N H3 



