2MgNH_{4}Po_{4} —­> Mg_{2}P_{2}O_{7} + 2NH_{3} + H_{2}O.

The precautions mentioned on pages 111 and 123 must be observed with great care during the ignition of this precipitate.  The danger here lies in a possible reduction of the phosphate by the carbon of the filter paper, or by the ammonia evolved, which may act as a reducing agent.  The phosphorus then attacks and injures a platinum crucible, and the determination is valueless.]

ANALYSIS OF LIMESTONE

Limestones vary widely in composition from a nearly pure marble through the dolomitic limestones, containing varying amounts of magnesium, to the impure varieties, which contain also ferrous and manganous carbonates and siliceous compounds in variable proportions.  Many other minerals may be inclosed in limestones in small quantities, and an exact qualitative analysis will often show the presence of sulphides or sulphates, phosphates, and titanates, and the alkali or even the heavy metals.  No attempt is made in the following procedures to provide a complete quantitative scheme which would take into account all of these constituents.  Such a scheme for a complete analysis of a limestone may be found in Bulletin No. 700 of the United States Geological Survey.  It is assumed that, for these practice determinations, a limestone is selected which contains only the more common constituents first enumerated above.

DETERMINATION OF MOISTURE

The determination of the amount of moisture in minerals or ores is often of great importance.  Ores which have been exposed to the weather during shipment may have absorbed enough moisture to appreciably affect the results of analysis.  Since it is essential that the seller and buyer should make their analyses upon comparable material, it is customary for each analyst to determine the moisture in the sample examined, and then to calculate the percentages of the various constituents with reference to a sample dried in the air, or at a temperature a little above 100 deg.C., which, unless the ore has undergone chemical change because of the wetting, should be the same before and after shipment.

Procedure.—­Spread 25 grams of the powdered sample on a weighed watch-glass; weigh to the nearest 10 milligrams only and heat at 105 deg.C.; weigh at intervals of an hour, after cooling in a desiccator, until the loss of weight after an hour’s heating does not exceed 10 milligrams.  It should be noted that a variation in weight of 10 milligrams in a total weight of 25 grams is no greater relatively than a variation of 0.1 milligram when the sample taken weighs 0.25 gram

DETERMINATION OF THE INSOLUBLE MATTER AND SILICA