1096 GEOLOGY OF OHIO. 
in general character, applied to something widely different is sure to 
lead to false results, unless each residue, precipitate and solution is 
examined, any phosphorus contained extracted and added to the total. 
Having obtained all the phosphorus in a clear nitric acid solution, it 
can be determined by precipitation with molybdic acid, and accurately 
weighed as magnesic pyrophosphate. The only uniform method, appli- 
eable to any iron, will be one that is accurate in all cases (and this must 
be proved for each case), and does not depend upon a supposed uniform 
loss, or balance of errors. Manganese was precipitated in the obtained 
filtrate, after separating the iron from the solution of pig-iron by sodic 
acetate, by bromine, as described under analysis of iron ores. 
FIRE-CLAYS. 
The general method was as follows: One gramme of the clay was 
fused with 5 or 6 times its weight of sodic carbonate, and the fused mass 
treated as in case of iron ores. The silica thus separated was weighed. 
The solution was then treated as in case of a limestone, taking care to 
Insure a complete separation of the alumina by long boiling till all the 
free ammonia was expelled. | 
The iron was determined volumetrically in a separate portion, 
fusing and dissolving as before. The titration was made with a dilute 
potassic permanganate solution, carefully standardized. 
The alkalies were determined by the J. Lawrence Smith method, 
using chloride of ammonia and carbonate of lime to decompose the 
clay. The alkalies being obtained in solution as chlorides, were evapor- 
ated to dryness after separating all lime and other bases, ignited to expel 
ammonia salts, then dissolved, and the solution filtered, and then again 
evaporated in a platinum dish, and the dry salt weighed. ‘The potassa 
was then separated with platinic chloride, and weighed as the double 
chloride of potash and platinum. The “indirect method” of determin- 
ing potassa and soda, while it did not “ work,” owing to the small 
amounts of soda present (this element only appearing in most cases in 
traces), led to the interesting discovery of lithia as a nearly constant 
element in the Ohio fire-clays. Too much chlorine having been found 
by volumetric determination in the dried salts obtained as above to 
supply the potassa and soda, the cause of the discrepancy was investi- 
gated, and finally Uithia was discovered with the spectroscope in quite 
(spectroscopically) considerable quantities. To put the question beyond 
