Field Experiments on the Fixation of Free Nitrogen. 653 
ness. Any one at all acquainted with this kind of work can 
appreciate the enormous value of these labours, the fruits of 
which will descend to a remote posterity, and will ever stand as 
classic models of all similar experimental work. 
To return to the plots. In order to reduce the fertility of 
these four plots, which in my scheme are numbered 21, 22, 23, 
and 24, the surface soil to a depth of about three inches, with 
its vegetable growth, was removed by paring. The surfaces 
were forked over rather deeply in January, 1888, and left to 
weather. In the following April, No. 23 was planted with 
barley. No. 24 with oats, and No. 22 with Italian rye grass. 
Abundance of seed was used, and the drills were pretty close 
together (seven inches), as it was expected the crops would be 
small ; the weeding was all done by hand. The remaining plot 
No. 21, was sown with broad red clover. 
Tlie results were very poor crops. The barley gave, grain 
and straw together, 10| cwt. per acre. The oat crop gave 
9 cwt. per acre. The grain of the barley and oat crops was not 
threshed out, but it would probably not have been more than 10 
bushels per acre. The Italian rye grass gave of green grass, in 
two cuttings, 41 cwt., which, dried into hay, would have been 
about 10 cwt. 
These crops will of course be regarded as exceedingly small, 
and they point clearly to the fact that the land had little or no 
“ condition ” so far as white corn-growing was concerned; in short, 
that its available (or nitric) nitrogen was at the very lowest ebb. 
I consider that this point is better demonstrated by the yields 
of these three gramineous crops than could have been done by an 
elaborate analysis in the chemical laboratory. This experiment 
proved, so far as Nos. 22, 23, and 24 were concerned, nothing as 
to the mineral constituents of the soil. These might have been 
present in superabundance, and yet have produced the same or 
very similar results in the absence of assimilable nitrogen. 
We now come to the broad red clover crop, on No. 21. The 
treatment of the land was the same as for the three other plots. 
The seed was sown in April, 1888. The crop was harvested on 
August 30, and niade into hay, the yield of which was at the rate of 
1 8 cwt. per acre. This was not a normal crop, but, as a plant of more 
than annual duration, it was likely to give a better result. The 
second year, unfortunately, it was by mistake top-dressed with a 
mixture of basic slag and kainit, and gave at two cuttings a 
yield of 9^ tons of green clover, equal to more than two tons of 
hay per acre. This top-dressing, of course, threw it completely 
out of the running. So far, therefore, this experiment affords 
us no means of knowing what available mineral plant-food the 
