ELISHA MITCHELL SCIEx\TIFIC SOCIETY. 23 



Bat the difference between the two finely pulverized samples after 

 eight (8) days of drying is only . 10 per cent. 



That the Am. Citrate should dissolve more Phos. acid from the 

 finely ground san pie is just what was to be expectei. When sufii- 

 cient Sulphuric acid is added to the crude Tri -calcium Phosphate 

 to render all of the Phosphoric acid Soluble in water reversion 

 does not appear to proceed as rapidly as when there is present some 

 of the original undecomposed phosphate. But in this casf^, when 

 using Charleston Rock, it is very difiicult to obtain a product which 

 will dry in a reasonable time without the aid of artificial heat, or 

 some carbonate as a dryer. Using artificial heat there is great 

 danger of hastening reversion, and the same is true if some chemi- 

 cal "dryer" is used, to say nothing of the reduction of the content 

 of total Phosphoric acid in this latter case. 



There is one point to which I wish to direct especial attenti- n, 

 and that is what I conceive to be the necessity for fine grinding of 

 the ancdytical sa7nx)le. Plants derive their food from the soil in 

 solution*, and in these solutions the food is in a state of almost in- 

 conceivable fineness. 



In estimating the value by chemical analysis of any plant food, 

 we should, as far as practical, approximate to the degree of fine- 

 ness to which the food must be reduced before the plant can use it. 



Other things being equal, the finer we grind analytical samples be- 

 fore acting upon them with chemical reagents which in a greater or 

 less degree represent the action of the soil the nearer do we approach 

 to the methods of nature. 



Laboratory of Navassa Guano Company, \ 

 Wilmington, N. C, Dec. 6tli, 1883. / 



ANALYSIS OF CHAPEL HILL WELL-WATERS. 



E. A. DeSCHWEINITZ, A. B. 



The sampl s examin d were drawn during the months of October 

 and November, 1883. The methods of analysis given in Cairns Avere 

 followed, as a rule. The oxidizable organic matter was roughly de- 

 termined by titration with potassium permanganate. The results 



