FIELD EXPERIMENTS WITH FERTILIZERS. 1 37 



increase the proportion of protein relatively more than it does 

 the yields of dry matter. The comparative yields show, too, 

 the great advantage in using nitrogenous fertilizers for the 

 purpose of increasing the feeding value of the crop. B3' their 

 use both the yields of total crop and the proportions of nitrogen 

 and of protein in the crop are increased. 



EXPERIMENTS WITH COW PEAS. 



By reference to the diagram on page 124 it may be seen that 

 the two series of sections adjoining each other and designated 

 as C and D were used for the experiments with cow peas. The 

 results obtained on the corresponding sections have been com- 

 bined and the data reported as if obtained from sections one 

 twenty-fifth acre each. The Whippoorwill variety of cow peas 

 was grpwn. As most varieties of this crop do not mature seed 

 in this climate, new seed must be obtained each 3'ear from fur- 

 ther south. The crop is harvested before or near the time of 

 blossoming, and is used for fodder. 



The amounts and cost of fertilizers per acre, the yields of 

 crop per section as harvested and the calculated 34elds per acre, 

 and the increase in yield on the fertilized plots over the average 

 of the yields from the two plots where no fertilizer was used, 

 are given in Table 70. The percentages and amounts of dry 

 matter in the crop as harvested, and the percentages and 

 amounts of protein in the dry matter, which are usually re- 

 ported, are here omitted. Samples for analysis were taken as 

 usual when the crop of 1901 was harvested, but because the 

 heater for operating the large drier was not in order the sam- 

 ples could not be dried by heating; instead of this they were 

 spread out thinly in a large room, but owing to continued wet 

 weather they moulded seriously before they dried, so that it 

 was deemed best not to analyse them. 



The amounts of nitrogen in the fertilizer and the total yields 

 of the crop. — From the results given in Table 70 it will be seen 

 that the yields per acre as calculated from the data for the dif- 

 ferent sections are very irregular. On the whole, however, 

 the indication is that the nitrogenous fertilizer had very little 

 effect on the j-ields. The best yield of any section was obtained 

 from one of the plots (6/^) supplied with mineral fertilizers 

 only, while the yield on the other mineral plot (6a) was equal 



