172 
Journal of Agricultural Research 
VoL XXV, Na 4 
give much information on this point, since many of the experiments were 
performed in quartz sand and in other experiments the quantities of lime 
applied bore no particular relation to the lime requirements of the soils. 
Also, very little work has been done with a view to ascertaining the rate 
at which the different phosphates gain or lose in efficiency by remaining 
in different soils. This is a question of considerable moment in the fertili¬ 
zation of long-time crops, such as fruit trees, sugar cane, and pineapples. 
The experiments reported in this paper were conducted to determine 
the relative efficiencies of acid phosphate, double superphosphate, basic 
slag, bone meal, and finely ground rock phosphate when used under differ¬ 
ent conditions in eight types of Porto Rican soils. The results are of 
general interest, however, in showing quantitatively how the various 
phosphates are affected by remaining in different soils and by applications 
of lime which are sufficient to satisfy the “ lime requirements ” of the soils. 
EXPERIMENTAL METHODS 
The relative efficiencies of the phosphates were calculated, not from 
the magnitudes of the increases produced by equal quantities of phos¬ 
phoric acid, but from the relative quantities of phosphoric acid required 
to produce the same increase in yield. For example, if 2 gm. of phos¬ 
phoric acid from floats produced the same increase as 0.5 gm. of phos¬ 
phoric acid from acid phosphate, the floats phosphoric acid was con¬ 
sidered as being 2 5 per cent as efficient as the acid phosphate phosphoric 
acid. In order to follow this method of comparison, it was necessary 
to have in each experiment a series of pots receiving increasing quantities 
of phosphoric acid from the standard material, acid phosphate. A curve 
was plotted from the yields of these pots to show the extent to which 
growth would be increased by any quantity of acid phosphate. 3 
As pointed out in another publication, (< 5 ) it is believed that this method 
of comparison is more likely to give accurate results than is the usual 
method because it is not only not based on any assumption concerning 
the law of minimum, but it is accurate, irrespective of how growth 
increases with increasing quantities of the elements in minimum; and 
it does not necessitate an analysis of the crop. 
Basic slag, bone meal, and floats or finely ground rock phosphates, 
were compared on the basis of their total phosphoric acid content, while 
acid phosphate and double superphosphate were compared on the basis 
of their “available’ 1 or ammonium-citrate-soluble, phosphoric acid. The 
efficiencies of the various phosphates are expressed as compared with 
the efficiency shown by acid phosphate when the latter was applied 
in the unlimed soil immediately before planting was done. 4 
In the larger experiments the differently treated pots were replicated 
five times or more. The probable error of the average result for each 
treatment was therefore fairly low—less in most cases than 4 per cent 
of the yield. The regularity of the curves showing the increases in 
growth produced by increasing applications of acid phosphate also con¬ 
firmed the accuracy of the results. 
Millet (Setaria italica ), which readily responds to phosphoric acid, 
was used as the test crop. As soon as the seeds had matured and the 
* A small preliminary test was conducted with each kind of soil to determine approximately the 
maximum quantity of phosphoric acid to which the crop would respond. 
* In Experiment VII the yields produced by add phosphate in the limed soil were taken as the standard 
for comparison, since in this soil add phosphate was very ineffective without lime. 
