the seed.

Finally, we see from experiments 11 and 12 that in all the foregoing
cases it was a limited supply of _nitrogen_ that limited the crop; for,
on adding Peruvian guano, which could only act by this element (its
other ingredients, phosphates of lime and potash, being abundantly
supplied in the ashes), the yield was carried up to 53.78 grammes, or
35-1/2 times the weight of the seed, and 13 times the weight of the crop
obtained from the unmixed peat.

5.--_The Examination of Peat (muck and marsh-mud) with reference to its
Agricultural Value._

Since, as we are forced to conclude, the variations in the composition
of peat stand in no recognizable relations to differences of appearance,
it is only possible to ascertain the value of any given specimen by
actual trial or by chemical investigation.

The method _by practical trial_ is usually the cheaper and more
satisfactory of the two, though a half year or more is needful to gain
the desired information.

It is sufficient to apply to small measured plots of ground, each say
two rods square, known quantities of the fresh, the weathered, and the
composted peat in order, by comparison of the growth and _weight_ of the
crop, to decide the question of their value.

Peat and its composts are usually applied at rates ranging from 20 to 40
wagon or cart loads per acre. There being 160 square rods in the acre,
the quantity proper to a plot of two rods square (= four square rods,)
would be one half to one load.

The composts with stable manure and lime, or salt and lime mixture, are
those which, in general, it would be best to experiment with. From the
effects of the stable manure compost, could be inferred with safety the
value of any compost, of which animal manure is an essential ingredient.

One great advantage of the practical trial on the small scale is, that
the adaptation of the peat or of the compost to the _peculiarities of
the soil_, is decided beyond a question.

It must be borne in mind, however, that the results of experiments can
only be relied upon, when the plots are accurately measured, when the
peat, etc., are applied in known quantities, and when the crops are
separately harvested and carefully weighed.

If experiments are made upon grass or clover, the gravest errors may
arise by drawing conclusions from the appearance of the standing crop.
Experience has shown that two clover crops, gathered from contiguous
plots differently manured, may s