ELISIIA MITCHELL SCIENTIFIC SOCIETY. 55 



water, and calculate as above (as with tea), or you may use 1, 2, 

 3 or 4 drops of the U!idilutcd infusion in the first glass and 

 make the calculation thus, /. e.: As the number of drops of 

 infusion used is to the number of drops "tannin solution" used 

 (to correspond), so is 5 to the answer— thus, suppose two drops 

 infusion were used and the corre3[)()nding tumbler contained fif- 

 teen drops tannin solution — -2 : 15 : : 5, answer 37.5 per cent. 



The object in diluting the infusions is because the infusion 

 glass may be of too deep a blue shade. It is better that it should 

 just produce a light blue. 



The tumblers must be perfectly clear and clean. 



The "iron mixture," "tannin solution" and infusion must 

 be fresldy prepared and not exposed to the rays of the sun. 



The water used must be free of iron and tannin. 



The results are necessarily in terras for commercial gallotannic 

 acid, and not for those of pure tannin, or of the particular tannin 

 in the material assayed. 



TEST FOR IRON. 



A solution of neutral sulphite of soda containing a little 

 pyrogallic acid has been proposed as a test for copper. A few 

 drops of it mixed with a dilute solution of a salt of copper 

 produces a red color similar to that which is developed by the 

 addition of sulphocyauide of potash to a solution of a persalt of 

 iron. The test is much more delicate for iron, as the followino; 

 experiment will show: 



Dissolve 0.7 gram ammonia ferrous sulphate (=0.1 gram iron) 

 in a liter of water; it will be 1 part in 10,000. To 10 cc. 

 of this solution add water to make 100 cc. ; this will be 1 to 

 100,000. Dilute some of this by adding four times its bulk of 

 water; it will then be 1 in 500,000. 



Make a saturated solution of sodium sulphite, and separately 

 a solution of pyrogallic acid 0.5 gram in 50 cc. water. Put 

 some of the iron solution in a wine-glass, add 4 drops of the 



