N. H. Agr. Experiment Station [Bulletin 306 



Table I. Three-year sumnmry of fertilizer treatments on 



Whenal farm plots. 



While the cheek plots were untreated so far as annual top-dressings 

 were concerned, they received the same basic fertilizer application at 

 seeding as the others. 



The response for nitrogen is more marked than for the phosphorus- 

 potash treatment. The increase obtained for 200 pounds nitrate of 

 soda was 1,462 pounds of cured hay, or 731 pounds for each hundred- 

 weight of nitrate of soda applied. AVith nitrate of soda currently 

 quoted at about $2.00 per hundredweight, the cost per ton of the in- 

 creased hay produced with this substance alone was $5.48. 



At current quotations the superphosphate and potash used in the PK 

 treatment will cost about $2.87. On the basis of the returns secured 

 with these materials the hay increase produced would entail an expendi- 

 ture of $7.88 per ton. When the same calculations are made for the 

 NPK treatment it is found that the hay produced from its use cost $6.36 

 per ton. These calculations do not include the cost of spreading the 

 fertilizer nor of harvesting the increase in the crop due to fertilization. 

 Nor do they take into account any possible increase in the nutrient or 

 mineral content of the hay due to its having been fertilized. 



Mathes Field Experiment 



A more elaborate top-dressing experiment was laid out in 1933 on a 

 field of the University farm in whicli nitrogen cari-iers have been com- 

 pared with each other and with complete fertilizers, and nitrogen has 

 been applied at different dates and at different rates. 



Small plots were used in this test and the material harvested from 

 each plot was transported to a drier where it was reduced to an air-dry 

 condition. This method eliminates any inaccuracies due to unequal 

 drying of the cured hay. Harvests were taken annually on June 10 

 and the plots, sixty in all. were harvosted on the same day. Each treat- 

 ment was repeated four times and tliere were twelve untreated plots in 

 all. A sample was secured from each plot. This sample was ground 

 and then analyzed for its protein content. 



The field on which this experiment was localed was acciuircd by the 

 University in 1921. The portion on which the plots were established was 

 plowed, heavily manured, and .seeded to mixed hay in 1923. 



The soil is a clay of the Suflfield series, a soil too heavy for .short rota- 

 tions, but one in which grass hay plants live for a long period and yield 

 quite well. Lack of fertilization during the period from 1923 to 1933 

 may account in large measure for the excellent responses, shown in the 

 data, that have been secured from complete fertilizers. 



