The cobalt salt under examination is mixed with an excess of sodium hydroxide and then treated with a solution of sodium hypochlorite.

Both of these reagents are added in excess, as the complete oxidation of the two metals is of primary importance. The solution is brought to a boil and the black precipitate, consisting of the two sesquioxides, after settling, is filtered off and is then treated with a solution of ammonia (1 : 3).

On applying heat the nickelic oxide is reduced completely, leaving the cobaltic oxide unchanged; this is filtered off and brought in contact with ferrous ammonium sulphate and sulphuric acid, when the ferrous salt will be oxidized at the expense of the cobaltic oxide.

The unchanged ferrous salt is then determined by titration with potassium permanganate.

On working with a solution containing 0.1240 gram cobalt and 0.1263 gram nickel, the resulting sesquioxides were treated with dilute ammonia and boiled for a short period; it was then filtered and treated with ferrous ammonium sulphate and titrated as recommended, with the result that there was found:

With 0.0496 gram cobalt and o. 1263 gram nickel, following the

With 0.1240 gram cobalt and 0.0506 gram nickel, there re

But, on making up an ammoniacal solution of ammonia (sp. gr. 0.96), with three parts water, as recommended, and boiling for a longer period, the results were more concordant. With a solution containing 0.0915 gram cobalt and 0.1263 gram nickel, there was found:

On lowering the nickel content to 0.0550 gram and submitting the oxides to the action of boiling ammonia for one hour, the results continued high.

When the period of boiling was extended to two hours, 0.0915 gram cobalt, in the presence of 0.1263 gram nickel, gave:

But when the nickel content was lowered to 0.0505 gram and the solution was boiled for two hours, the results were high:

These results show conclusively that the weakness of the method is in the reduction with ammonia. The proportion of the two metals is also shown to have an effect on the determination: the proportion of 2 cobalt to 1 nickel seemed especially

unfavorable.

IV. DONATH'S METHOD.

In the Berichte, 1879, p. 1868, appears a method devised by Donath. It is really proposed as a modification of the Fleischer method. The instability of the sodium hypochlorite is a bar to its general acceptance, as is also the reduction by ammonia.

As a modification, Donath suggests the oxidation by iodine in a sodium hydrate solution, whereby the cobalt is completely converted into the sesquioxide state, while the nickel remains as

nickelous oxide. The precipitate may then be filtered off and the sesquioxide brought into a distilling apparatus along with hydrochloric acid, when the following reaction takes place :

Co,O,+6 HC1 = CoCl, + 3 H2O + Cl2.

The liberated chlorine is conducted into a potassium iodide solution and the liberated iodine is titrated by sodium hyposulphite. From the iodine found the cobalt can be readily calculated. On oxidizing a solution containing 0.1263 gram nickel and 0.1240 gram cobalt with thrice sublimed iodine, the results were always high, as the following series will show :

But on using chemically pure iodine the results were much

0.1388

111.93

0.1607

129.60

When the nickel content was reduced to 0.0506 gram with 0.1240 gram cobalt, the results were as follows:

From these results it can be seen that the purity of the iodine is of prime importance, otherwise part of the nickel will be oxidized. The method may be considered as fairly accurate.

V. VON REIS' METHOD.

In the Zeitschrift für angewandte Chemie, 1890, appears an article by M. A. von Reis and F. Wiggert on the determination of cobalt in the volumetric way.

The cobalt solution is treated with an emulsion of zinc oxide and brought to boiling, when a known quantity of a standardized solution of potassium permanganate is added.

The oxidized cobalt falls to the bottom and allows of the backtitration of the potassium permanganate by ferrous ammonium sulphate.

To all appearances this is by far the most satisfactory method yet proposed. Its simplicity recommends it.

CONCLUSION.

After careful repetition of these methods, making varying conditions wherever deemed advisable, one is justified in concluding that none of them possess the degree of accuracy required in any trustworthy determination of cobalt. A good volumetric method for this purpose still remains to be devised.

UNIVERSITY OF PENNSYLVANIA.

T

RECENT WORK IN ENGLAND ON THE PURIFICA-
TION OF SEWAGE.'

By LEONARD P. KINNICUTT.

Received January 24, 1898.

HERE are few questions connected with municipal government which cause as much trouble as the disposal of sewage in such a manner as to prevent any cause for complaint, and it is certainly not far from the truth to say that, at the present time, there is no large inland city in the world that has succeeded in satisfactorily disposing of the refuse matter that passes through the sewers.

1 Read at the Washington Meeting, December 29, 1897.

The three methods used for the purification of sewage are: irrigation, or sewage farming; chemical precipitation; and intermittent filtration.

The first two methods have been tried by many large cities, and opinions differ as to their comparative merits. The history of sewage disposal is evidence of this difference of opinion.

There are in Great Britain forty-three cities which have a population of over 70,000. Direct information from personal investigation shows that, in 1897, twenty-one of these cities discharged their sewage practically untreated into tidal rivers, or into the sea Liverpool, Edinburgh, Dublin, Belfast, Bristol, Newcastle, Hull, Portsmouth, Dundee, Cardiff, Oldham, Sunderland, Aberdeen, Brighton, Birkenhead, Swansea, Gateshead, Plymouth, South Shields, Middleborough, Cork, and Stockport; nine used chemical precipitation as a means of purification: London, Glasgow, Leeds, Sheffield, Bradford, Salford, Bolton, Huddesfield, and Southampton; two, chemical precipitation, allowing the effluent to pass over a small area of land: Manchester, 30 acres, and Burmley, 30 acres ; five, broad irrigation : Leicester, Blackburn, Norwich, Preston, and Walsall; two, broad irrigation preceded by chemical precipitation: Birmingham and Wolverhampton; one used chemical precipitation for a portion of the sewage and broad irrigation for the remainder : Rochdale, chemical precipitation with 250,000 gallons and broad irrigation with 1,000,000 gallons. Of the German cities containing over 70,000 inhabitants, only Berlin, Breslau, Madgeburg, Frankfort, Danzig, Charlottenburg, Braunschweig, Halle, and Essen treat the sewage. Frankfort, Halle, and Essen by chemical precipitation, the other cities by broad irrigation, Berlin using for the purpose 14,473 acres and having in reserve 6,737 acres.

In all these cities there is great difficulty in obtaining satisfactory results. Those cities using broad irrigation find that more and more land must be added to the so-called sewage farm to prevent the drainage from causing a nuisance, and in those cities where the chemical precipitation method is employed, there is no doubt that there has been great difficulty in obtaining sufficient purification, especially when the surrounding district is extensively occupied, and where there is no large watercourse to carry away the purified effluent. The question, there