NOTES UPON WATER ANALYSES BY T H E AMMONIA METHOD WITH SOME N E W APPARATUS.*
B Y A. A . B R E N E M A N .
Wanklyn's ammonia method,of water analysis, in spite of its imperfections, is still the process most widely used in this country. I t is the simplest: and most practicable of all combustion methods, wet or dry, that have been applied to water analysis, and although it shares with all processes of this class used in sanitary work, the doubt as to interpretation of its results, its general use among chemists indicates its superior adaptability to such work. The process as now used is essentislly the original one, but experience has shown that the interpretation of results must be guided by familiarity with the method and by a knowledge of the class to which the water in question belongs. Also, there is now a marked tendency to prescribe more closely than Wanklyn has done the minute details of the process. The variability of naturai water with respect to the indications which they yield under Wanklyn's process, as to time of distillation, concentration of reagents and rate of attack of oxidable matter, have necessitated the adoption of standard methods as the nearest approach to a means by which the results obtained froin different waters, and by different observers, could be made in any sense corn parable. I have shown in a previous paper (This Joz~wzd,Vol. 8, No. 9) the points in which such prescription of detail should be insisted upon and have suggested a standard method for general use. The previous work of Mallet, Smart, and others, had suggested many similar points of detail requiring attention. The Report of the Committee of the Chemical Section of the American Association for the Advancement of Science for lSS9 -_
*Read at the Newport Meeting.
upon the methods of water analysis adopts many of these different suggestions. With regard t o the quantity of water to be used tlic quantity is left optional, as to 300 or 100 c.c., i l l this report. 111 common with many other chemists I should have been glad to sei3 it recommendation as to the exclusive use of the 100 C.C. mrtlio~l and of 10 c.c. nesslerizing tubes :ts represciiting a neater n r i t l more accurate method of‘ c*oiiilncting tlic aii:ilysis. ‘I’lie tis(’ of tlie smaller quantity of water j s 110 exception to t h e gt~1ioritIi)i’c~fcrerictc i n chemical analysis f o r small saniples w l i o i ~acc*iirttc~yis n o t i n i paired by their use. In the preseiit c;tw there are ti) IJC, ofher1 the advantages of less cumbrous apl):tr:ttiis, :L 1insteiiiiig of the process and itii increase in : i c ~ u r i i c yljy the iise of tht: xtii:illc~rtullcbs with their greater de1)tIi for a givc,il volumc: of the Iiqriid subniitted to t h e colorimetric test. Anotlicr detail upon which I have insisted, its ~icce~s:iry i n order to render resnl ts obtaiiiecl by the K n n k l y n process coriilxtrablc., as nearly :is possible.. f o r wttc1.s of all cl:iskes, is the matter of carrythe distillntioii t o :i vcry small rc~sic1ii:tlvolume i n tlie retort, SO t h a t tlic gretLtest streiigtli of t h e ~ ~ e i ~ i i i ~ ~ i i ~ and : t i i i tait ehili and the strongcst :itt:ick upon ositl:tblc mntter ~liitllt)c coiiceiitratrel upoii the oi~g:tnic matter ~ 1 i i ~ : l has i longest rrsistctl that, itttack ; i i i other \voids the W I iiig of tlie distillatioii to a imiiit I I C elryness. ~ ~ This itgxin is much more ensily done in a small tliaii i i i a large retort, f o r obvious rcasons. Fnrther. t h e accurate measurement of all Iiqiiids put into the retort permits a control o i tlic clistilhtion by measnremcnt of t l e 10 c.c. portioiis wliich :LIT taken off as distilltites during the operation, 01’ of tlic 65 C.C. iiui.tioii takcn off i n the preliminary process of cleansing the retort :is already describetl. (Zoc. cil.) T h e degree of coiidensation iiiag thus be judged of and the volume of the residual liquid in tire retort :it the close of tho operation may be known within 2-3 C.C. a t most. I n practice I leave less than 5 C.C. of liquid in thcs retort at tlit: ciose, yet accidents from running d r y o r spurting arc almost u 11known. The objection that,modification of the old metliotl will prevent comparison of t h e results with those previously obtainrd, is an objectioii that, in some sense, mects every attempt a t improvement of an analytical proccss and cannot be regarded as iiisnperable.
