IIlPROVED IlETHODS OF WATER ANALYSIS. BY
IRVISG A. RACHHAN.
R e c e ~ v e dJ , m u a r y
2,
1895.
E have in the South Atlantic States incontestable evidence that the true source of malarial disturbance is to be found in the water, and not in tlie air, as was formerly supposed. T h e introduction of deep-seated artesian wells has very materially decreased this trouble, and has rendered localities that were notoriously unhealthy, perfectly healthy. T h e examination into the exact difference between the malariaproducing waters and those proof against it, has had the writer's attention for over two years, and the work herein detailed is a portion of the preliminary work looking into the best methods of analysis. My experience with the Wanklyn process as ordinarily carried out with the usual apparatus, so impressed me with its crudeness, that for a long time I have had little confidence in the results I obtained. T h e loss by imperfect condensation, the crude way of adding the permanganate solutions, and the openair coiltamination of the (listillate are by no means in keeping with accurate work. The loss by imperfect condensation i s very niuch greater than we permit ourseives to believe, and with the original Wanklyn process' is a very dificult matter to control ; the loss is variable, and is dependent upon the rate of distillation and tlie efficiency of the cooling apparatus. After inaking a long series of experiments upon every character of material liable to be found in water. I am satisfied that a rate exceetliiig fifty cc. As a modification of in .fifteen minutes is accompanied b y lo tlie Wanklyn process, Mallet' proposes to keep the original voluine of liquid in the retort constant b) the systematic addition of ammonia-free, distilled water, 'in ortier to prolong the action of the perinaiigaiiate ; this is objectionable, inasmuch as the operation is endangered by a constant soiirce of error i n the addition of aninioiiia-free water, which is difficult to obtain ill so large a quaiitit!,. 1 \EI~ii!.lyii
and C h a p m a n Water Analysis. (6th e d . ) , 1885, p p . 38-40. Hoard of I leaith. I X X Z . 1). Z I L )
2 licktort Satiorial
I M P R O V E D METHODS OF IV.4TER AXALYSIS.
297
I n my own work I had frequently noticed that the evolution of ammonia very often increased as the liquid in the retort became niore concentrated. Acting on this observation it occurred to me that by the action on a sinaller arid limited quantity of water by the full strength of the pernianganate solution and supplying the water under examination at about the rate of distillation, very much better results could be obtained ; this I have now confirmed by a great many comparative analyses. T h e apparatus used iS shown in Fig. I , and consists of distillation flask ,4, of one liter capacity, into which is fitted by a grouiid joint the remainder of the apparatus (this is the only
Q
298
IRVING A. BACHMAN.
of the bottom of the distillation flask : the larger burette C , has a capacity of z j o cc. and its tube projects to within one-half inch of the bottom. T h e cooling apparatus D, is a closely coiled glass worm, with a long projection at the lower extremity, to which is attached the distillate vessel E , by a soft rubber stopper, which also has tube for connection with the %‘ill and Varrentrapp absorption bull). T h e apparatus is constructed to work under greatly reduced pressure, and the connection with distillation flask, as well as the stop-cocks, are accurately ground tt, remain tight without a lubricant. T h e apparatus is mounted as shown, raised and lowered into the bath by nieaiis of a thumb-screw, as indicated. T h e Ivhole is very easily managed, and is very serviceable. I have made over a thousand tleterniinatioiis with mine, arid there is no reason why it will not continue to do good service for a very long time to come. T h e heating of the apparatus is effected b y setting the flask into a fixed saline-bath, with a covered top as shown, and heated froin below by a two-pipe gas-burner. Mode of Proccdu Y P .-After t horouglil y r i iisi ng the apparatus with ammonia-free distilled water, j O O cc. of the water urider examination are put into the flask arid the apparatus connected and lowered into the bath, the distillate flask with it-ill and Varreiitrapp bulb, the latter previously charged with Nessler solution, are now attached and connected with the pump.’ T h e partial vacuum is now established gradually, the sodium carbonate solution added by way of the smaller burette, and the distillation started and continued until zoo cc. have been distilled over, which is then nesslerized for free aninionia. T h e vacuuiii tube is now disengaged and the stop-cock on large burette opened and about 2 5 0 cc. of the liquid in the distillation flask drawn u p arid the fifty cc. of alkaline permanganate solution, previously poured into the smaller tube, is now allowed to run in after t h e distillate flask has again been attached. 1 Richard. is nil excellent device. b u t i t i:, irnportant to have a slnall check-valve in the tuhe to prevent the w a t e r being d r a w n into the apparatlls when the water pressure i s suddenly diminished. A w r y simple valve is made of oiled silk opening otlly toward the punig.
IMPROVED METHODS OF WATER ANALYSIS.
299
W e now have fifty cc. of alkaline permanganate acting on fifty cc. of water, and after, say thirty to forty cc. have been distilled over, the water in the large tube is allowed to drop in a t the same rate a s the distillation, which must not exceed fifty cc. in fifteen minutes. [ T h e rate of distillation, as recommended by the Chemical Section of the American Association for the Advancement of Science, i. e . , fifty cc. in ten minutes, is too rapid, and is very frequently accompanied by loss by imperfect condensation ; this is very readily observed in my apparatus by discoloration of the Nessler solution in bulb.] T h e distillation is continued until 250 to 300 cc. have been distilled over, which is then nesslerized. T h e following notes from my laboratory work will show the general plan of the work : One gram urea was dissolved in one lifer ammonia-free water. I . T e n mgnis. urea in 300 cc. of ammonia-free water, 250 cc. drawn u p into tube, fifty cc. pernianganate solution added, bath 87O C., cooling water 28' C., pressure s i x and a half inches, distillation a t rate of 150 cc. per hour, 310 cc. distilled over with slight discoloration of Nessler bulb. 30.2 : 2 . 0 : : 310 : 2.053 mgms. ammonia, or 20.53 per cent. which is 36.2 per cent. on ammonia obtainable. 2 . Five mgnis. urea in 300 cc. ammonia-free water treated a s above--297.4 cc. distilled over, no discoloration of Nessler bulb. 28.8: 1.0 : : 297.4 : 1.032 nigins. ammonia, or 20.64 per cent. which is 36.4 per cent. on amiiioiiia obtainable. 3. T e n nigms. urea in 300 cc. water, 2 5 0 drawn up, fifty cc. permanganate solution added, bath 85' C., cooling water 27' C., pressure five inches, rate of distillation, 100 cc. per hour, 290 cc. distilled over, no discoloration of Nessler bulb. 25. I : 2 . 0 : : 290.0 : 2.310 nignis. ammonia, or 23. IO per cent. on 40.7 per cent. 011 ammonia obtainable. 4. Teii nignis. urea in 500 cc. water, ten cc. strong solution sodium carbonate added, bath 87' C., cooliiig water 28' C., pressure five and a half inches. 2 1 0 cc. distilled over as free ammonia, which yielded 0.835 mgms. aninionia or 8.35 per cent. 250 cc. were then drawn up, fifty cc. permanganate solution added, and distilled as before.
300
IRVING
A\.
13.LCHX.LS.
305 cc. were distilled over, which yielded 2.296 iiigms. aiiiiiionia, or 22.96 per ceiit. Total ammonia 8.sj per cent. free, plus 22.96 per ceilt. albuminoid, or 3 I .3 I per cent. which is 5 j,2 per ceiit. 011 niiiiiioiiia obtainable. Rosaiiiline hydrochlorate yieliletl 86.02 per criit. ;9.10 '< Strychnine sulphate Urea-boiling five hours " 79.90 " Cyanuric acid 61.10 Quiiiiiie sulphate ;:.(y Morphine sulphate " 91.11 "
011
ammonia obtainable.
"
"
Tlie apparatus, Fig. I , is also used for the Kjeldahl process. This process was first described for use in general water analysis by DroLvii and blartin,' arid is iiow very generally used in place of the Ii-anklyri process. I har-e modified the Kjeldahl process so as to insure more perfect decomposition, and have obtained most gratifying results with the most obstinate organic compounds. Pyoct>ss.-po cc. of the water under exainiiiation is put into the distill?tion flask of apparatus alreadj. described ( F i g . I ) and the wliole coiiiiectecl up ; ten cc. specially prepared concentrated sulpliuric acid is allon.ed to drop in slowly by way of the smaller burette, the nliole is then allowed to digest at 8 j " C . , under diiniiiislieti pressure for, say one hour, and then brought to a boil, aiitl so continued until tlie contents are colorless. T h e acid is tlieii neutralized and the solution rendered strongly alkaline b y the addition of fifty cc. strong solution of aininoiiia-free, sodiuiii hydroxide through t!ie siiialler tube. 2 j0 cc. of the coiiteiits are theii drawn up into the larger tube and fifty cc. of perinanganate solution added through the siiialler tube, and the distillation carried 011 as already described in the iiiodificatioiis of the M'anklyn process, and nesslerized a s before. T h e modified process coiiibiiies the good features of t h e U'anklyn and Kjeldahl, and is easily carried out with the apparatus described. T h e original sources of error are now practically eliminated. The following are soiiie results upon some of tlie substances 1
'I'echiiology Q i i n r t e r l y .
2 , j.
IMPROVED METHODS O F WATER ANALYSIS.
30'
that with the Wanklyn process yield only a very small percentage of the ammonia. I have made no comparative tests with the original Wanklyn process, and quote Mallet's :' Mine.
Mallett, by Wanklyn's Process.
Rosaniline hydrochlorate 96.9 per cent. 23.8 per cent of theoretical amount. T y a n u r i c acid
