62 BUREAU OF AMERICAN ETHNOLOGY [bull. 36 



loss in weight recorded as volatile matter; this comprises firmly 

 combined ^Yater, carbon dioxide, and orijanic substances. 



" The sample was then dissolved in dilute nitric acid, and the solu- 

 tion filtered, the residue being ignited and weighed and examined 

 microscopically. In every case it proved to be made up of quartz- 

 sand and clay ; and, as such material is clearly a mechanical admix- 

 ture with the bone, it was deducted in each case from the total weight 

 of sample taken, so that the percentages of material calculated would 

 represent material combined in the bone substance and the results on 

 the several samples would be directly comparable. 



" The nitric acid solution was heated to boiling, made slightly alka- 

 line with ammonia — a small amoimt of mixed phosphates being 

 thereby precipitated — and an excess of acetic acid was added. Cal- 

 cium phosphate then went completely into solution, while the phos- 

 phates of aluminium and iron remained insoluble, and were filtered 

 out, ignited, and weighed. The amount of iron was so small that it 

 was not separately determined. When the precipitate was large in 

 amount it was fused with sodium carbonate, dissolved in nitric acid, 

 and the phosphate removed by ammonium molybdate, and deter- 

 mined as magnesium pyrophosphate; the phosphoric oxide found 

 was deducted from the total precipitate, and the remainder recorded 

 as ALOg (-t-FejOs). When the amount of precipitate was small, 

 the phosphoric oxide in it was calculated on the basis of the com- 

 pound AIPO4 ^^^ subtracted as before. 



" The acetic acid solution, from which aluminium had been re- 

 moved, was diluted to a definite volume and divided into two equal 

 parts. To one part ammonium oxalate was added and the calcium 

 oxalate precipitated was filtered out, ignited, and weighed as CaO. 

 To the filtrate from this ammonia was added and the magnesium- 

 ammonium phosphate precipitated was ignited and weighed as the 

 pyrophosphate, from which the MgO was calculated. Phosphoric 

 oxide was determined in the remaining half of the acetic acid solu- 

 tion by titration with a standard solution of uranium acetate; 

 the amount in the aluminium precipitate being added to give the 

 total P2O5. 



" For the determination of fluorine a fresh sample weighing 0.4 

 gram was dried, mixed with pulverized quartz, and heated to 200° 

 with specially concentrated sulphuric acid in a stream of dry air, 

 acid fumes carried over being removed by filtering the air through 

 dry asbestos, and the gases being collected by bubbling through water. 

 The hydrogen fluoride dissolved in the water was titrated with stand- 

 ard sodium hydroxide solution, and the percentage of fluorine present 

 in the bone thus obtained. 



" The specific gravities were determined on the powdered bone, 

 using the picnometer, at about 20°. 



