REPORT OF THE DIRECTOR. 9 



the yield or the composition. The effect of the nitrogenous 

 fertihzers upon the yields of the crops has long been known. 

 The principle which has been less thoroughly understood, and 

 which these experiments are bringing out more clearly, is that 

 not only the total amount of the crop, but also the percentage 

 of its costliest ingredient, protein, may be increased. In this 

 respect particularly the results of the experiments are of im- 

 mense practical importance to farmers, in indicating how the 

 production of protein on the farm may be increased. 



Pot experime7its with fertilizers . — Experiments similar to the 

 special nitrogen experiments on the field plots have been car- 

 ried on for a few years on a small scale in such a way that 

 moisture and other external conditions affecting the growth of 

 the plants might be more completely under control. In these 

 experiments the plants are grown in pots filled with soil quite 

 uniformly mixed, and treated with the same kinds of fertilizers 

 as are used in the plot experiments. The results of these 

 experiments accord with those of the special nitrogen field 

 experiments in indicating that the nitrogenous fertilizers are 

 valuable for increasing the quantity and protein content of 

 grasses and grains, but are of little value for the legumes. 



Experiment on soil ifnprovement. — The purpose of this exper- 

 iment is to study the economy of different methods of manuring 

 for restoring fertility to a soil lacking in organic matter and in 

 available nitrogen. The experiment was begun in 1899 and 

 has since been continued on the same plots. The results thus 

 far obtained indicate a high value for legumes like common red 

 clover when plowed under for green manuring. 



FEEDING EXPERIMENTS WITH DAIRY HERDS. 



Studies of rations fed to milch cows were carried on during 

 the past year as in preceding years, in cooperation with a num- 

 ber of farmers in different parts of the State. A report of the 

 studies of four herds is given by Prof. Phelps in the present 

 Report. The object of these experiments is to learn how 

 representative dairy farmers in Connecticut feed their cows, 

 compare the results obtained by their methods with results of 

 other methods elsewhere, and to suggest improvements wher- 

 ever they can be made. Besides serving for instruction to the 



