1 62 



MAIXE AGRICULTURAL EXPERIMENT STATION. 



There was little difference in the per cent of germination in 

 the two lots ; but the plants which were sub-watered were supe- 

 rior to the others from the beginning. At the time of harvest- 

 ing, the number of first-class roots on the sub-watered section 

 exceeded that on the surface watered section by i5% ; while the 

 average weight was 14.5% greater. This difference is plainly 

 shown in the accompanying figure. 



Second Trial: A second bench in the radish house was 

 divided into two sections, and the advantage of sub-watering 

 was demonstrated on a commercial scale. In this instance, the 

 arrangement for sub-watering was somewhat different than in 

 the previous one. The bottom and sides of the bed were coated 

 with Portland cement. On this was placed about two inches 

 of potsherds and broken brick, and then, after covering the brick 

 with some pieces of burlaps, the soil was put in place. As the 

 soil became dry, water was admitted through a pipe to the 

 stratum of potsherds. A glass indicator served to show the 

 height of the water. Each section of the bench was fourteen 

 feet eight inches long and twenty-eight inches wide. The seed 

 was planted in rows, eight inches apart, between rows of let- 

 tuce, — twenty-two rows in each lot. 



There was little dift'erence in germination of the two lots, but 

 very many more plants were lost by damping oft' on the sur- 

 face-watered section, — a fact that partly accounts for the 

 difference in yield at harvest time. 



The results at harvest were as follows : 



YIELD OF RADISHES FROM SUB-WATERED AND SURFACE- WATERED 

 BENCHES COMl'ARED. 



Treatment. 



^ 6 



3 



III 



3 -? 

 ■zZz 



Number of 

 roots, second 

 class and 

 culls.* 



1 1 



6 

 3 





28 

 13 



21-2 

 118 



242 



228 



Grams. 

 3,752 



1,498 



Grams. 

 2.590 



1,750 



4«.T 





34.1 









*The reason for the very high percentage of "second class and culls" is ex- 

 plained in the next table. 



