188 BULLETIN OF THE BUSSEY INSTITUTION. 
of my field experiments, published in the first volume of the “ Bussey 
Bulletin ” (see, for example, pages 86 and 136). Both oyster-shells in 
fine powder and oyster-shell lime did better fertilizing service in those 
experiments than the other kinds of lime that were employed in com- 
parison with them; but the oyster-shell lime was distinctly better than 
the unburnt shells, although the latter were extremely finely powdered. 
The following table, which gives the number of pounds of phosphoric 
acid, potash, nitrogen, and magnesia contained in a ton (2,000 lbs.) of 
the several kinds of shells examined in this laboratory, will perhaps 
present the economic side of the question a little more clearly than the 
previous table of percentage composition : — 
| | Contains (in round numbers) pounds of— 
A Ton (2000 Ibs.) of the 
Dry Shells of the -— Phosphoric Potash Nitrogen sss eae 
N. gO. 
Acid POs. K,0. 
Lobster (eh ore 624 794 314 
Common Crab. . . 71 39 
King Crab 
Oyster. 
oe 
Clam . 
Quahaug . 
Hen Clam 
Smaller Mussel 
Horse Mussel . 
—_ 
54 
$ 
Scallop 
Sea Snail 
Coh-t pt CoD oft pole! cops het fe 
Sho ca toh cope Robt sje en sbfoo 
Methods of Analysis. For the estimation of phosphoric acid, a quantity 
of the dry powder was dissolved in strong nitric acid, and the solution was 
evaporated to dryness, and heated to 110°, to render silica insoluble. In 
the filtrate from silica, the phosphoric acid was thrown down with molyb- 
date of ammonia, and estimated in the usual way. See page 24. 
For the estimation of potash, a quantity of the shell was ignited in a 
crucible, or roasted in a muffle to destroy organic matter, chloride of am- 
monium was mixed intimately with the calcined powder, in the proportion 
of one part of the chloride to about eight parts of the powder; and the 
process was proceeded with precisely as in Professor Lawrence Smith’s 
method of estimating potash in silicates. As much as 20 grms. of the shell 
were operated upon, i.e., ignited with the chloride, in most instances. 
