~ 
1884] 695 [Genth. 
The analysis of herderite presents great difficulties, and the following 
methods have been used in the different analyses : 
I. A portion of the material was slightly yellowish, and probably con- 
taminated with a trace of mica, 1.0334 grm. were slowly ignited toa 
bright red heat in a well covered platinum crucible, and lost 4.80 per cent ; 
after ignition, the lid and outside margin of the erucible were found 
coated with a film, which, when moistened with water, gave a strongly 
acid reaction. 
About equal weights of silica and the mineral were fused with about six 
parts of sodium carbonate for one hour. The well-fused mass, which was 
greenish from a minute quantity of manganese, was lixiviated with water 
and should now have contained all the phosphoric acid, and fluorine. ‘The 
silica in the solution was precipitated by ammonium carbonate, and from 
its filtrate the remaining silica by zinc oxide, dissolved in ammonium car- 
bonate and ammonia. From the filtrate of the zinc silicate after evapora- 
tion to dryness and dissolving in water, the greater portion of the sodium 
carbonate was neutralized with dilute nitric acid, then precipitated with 
silver nitrate, keeping the liquid slightly alkaline. In the filtrate the 
balance of the fluorine (which had not been expelled by ignition), was 
precipitated together with calcium carbonate by calcium chloride. The 
mixed precipitate was ignited, dissolved in acetic acid, evaporated to dry- 
ness, the calcium acetate dissolved out by water and the calcium fluoride 
determined. It gave 1.57 per cent which would give, with that driven off 
by ignition, about 6,4 per cent, 
The zine silicate precipitate was found to contain phosphoric acid. 
The portion insoluble in water was dissolved in a little hydrochloric 
acid and the clear solution precipitated by ammonia, and. this separation. 
twice repeated, and washed, until the filtrate was free from lime. The 
precipitate was ignited and weighed. It gave over 87 per cent and con- 
tained a large quantity of phosphoric acid. It was then re-dissolved in 
hydrochloric acid, the excess driven off by evaporation, then dissolved in 
water and enough sodium hydrate added to re-dissolve the greater portion 
of the precipitate. From the filtrate of the insqluble partion the phosphoric 
acid was precipitated by baryum hydrate as baryum phosphate ; in the fil- 
trate from this the excess of baryum was removed by sulphuric acid, and 
from the filtrate of the baryam sulphate, the glucina and alumina were 
precipitated with ammonia, The residue insoluble in sodium hydrate _ 
contained about 4.5 per cent of lime, nearly 8 per cent of phosphoric 
acid, a little iron and glucina and alumina, which latter were added to 
the ammonia precipitate. 
These were then re-dissolyed in as little acid as possible, precipitated by 
ammonia, and the separation of glucina from alumina attempted by that 
method, recommended by the highest authorities as the most perfect, 
namely, boiling these hydrates with a concentrated solution of am- 
monium chloride. The strength of the solution was kept so, that on 
PROC, AMER. PHILOS, 800. xxJ. 116, 4g. PRINTED DECEMBER 19, 1884, 
