306 ADDENDA. 



ard of the peimanganic solution is unsafe, for the reason that the com- 

 position of the salt is not constant, and that, consequently, metallic iron 

 or oxalic acid must be used. Fresenius also, gives the preference to the 

 use of metallic iron as the most accurate, although perliaps less conven- 

 ient, method; he gives tlie following directions for executing the proc- 

 ess. Weigh out about 0.3 grm. of the finest piano-forte wire, free from 

 rust, and add to it in the long-necked flask (p. 68) 30 c.c. of dilute sul- 

 phuric acid, and as much water, and proceed with the solution in the 

 current of carbonic acid, and the subsequent titration, as directed in the 

 case of the use of the ammonio-ferrous sulphate. Instead of £, 



T 

 in the proportion on page 69, substitute F x 0.997 ; this product is taken 

 as the weight of pure iron in the weight F of iron used, since the purest 

 wire contains about 0.3o|q of impurities; 



In the text (pp. 69, 154) the preference is gfven to a sulphuric-acid so- 

 lution of the ferrous salt for titration rather than a solution in hydro- 

 chloric acid. In an appendix to his Quantitative Analyse^ Fresenius states, 

 that the titration of a hydrochloric-acid solution is unreliable unless con- 

 ducted as follows. Make the volume of the solution up to a quarter of 

 a litre, add 50 c.c. of this solution to a considerable quantity of water 

 acidified with sulphuric acid, titrate this mixture with permanganate, 

 add to it 50 c.c. more of the ferrous solution, titrate again, and so on 

 with a third and fourth portion of the same solution; finally use the 

 last two results, the mean of which multiplied by five will give the 

 amount of permanganic solution that should be required for the whole 

 amount of the ferrous solution. 



The precipitate of ammonio-magnesic phosphate, in the esti- 

 mation OF magnesium, page 64, AND OF PHOSPHORIC ACID, PAGE 87. 



As the results of experiments by Kubel and by Kiesel (Fres. Zeitschrift 

 8, pp. 135, 164), it appears that the solubility of this precipitate in the 

 liquid in which the precipitation takes place, in the presence of con- 

 siderable amnionic chloride and a not too great excess of magnesic sul- 

 phate, is very nearly compensated for by the minute quantity of basic 

 magnesic sulphate, oi-of magnesia, that is precipitated at the same time; 

 hence the correction for the imperfect insolubility of the precipitated 

 phosphate seems to be unnecessary. 



Organic matter in water, page 273. R. Angus Smith, who is a 

 strong and earnest champion for the permanganate process, in the ex- 

 amination of water with respect to the pi-esence of hurtful organic 

 matter, uses a stronger solution than Kubel (see page 151). He adds 3 

 grms. of the salt to a litre of water, and gives the following directions 

 for the titration : 



To not less than a litre of the water add a drop of this permanganate 

 solution ; stir the liquid well, and wait until the color disappears, and 

 proceed in this manner as long as the color disappears quickly, say in a 

 minute or two ; genemlly, considerable permanency of color is obtained 



