18 UNIV. OF N. H. AGR. EXPERIMENT STATION [Bulletin 216 



T.^BLE VII. — Yields obtained by the use of Irish Cobbler seed cut from small and large tubers. 



The yield per acre of marketable tubers from the large seed was 271.38 

 bushels and from the small seed 237.14 bushels, or a difference in favor of the 

 large seed of 34.24 bushels per acre. 



In the case of the Green Mountain potato the following results were obtained: 



T.\BLE VIII. — Yields obtained by the use of Green Mountain seed cut from smill and large tubers. 



It is interesting to note that in the case of the Green Mountain and Cobbler 

 potatoes the large seed produced about an equal increase in gross yield per acre 

 over the small seed. On grading, however, the small Green Mountain seed 

 produced a higher percentage of marketable tubers than the large seed. The 

 large seed produced 227 bushels per acre of marketable tubers, the small pro- 

 duced 223.77 bu.shels, which gives a difference of only 3.23 bushels in favor of 

 the large seed. Since, however, small seed of good quality costs very nearly as 

 much as large seed, there is no reason to expect that its purchase will prove 

 profitable when it is used cut. 



Potato Seed Certification. 



During the year 1923, 55 acres of Green Mountain were inspected and 46.5 

 acres were passed for certification. The total gross yield of potatoes produced 

 was 15,182.75 bushels, and practically the entire crop was sold within the state. 



Soil Rejuvenation Studies. 



This project was started in 1920 by H. R. Kraybill (Agricultural Chemistry) 

 and M. G. Eastman (Agronomy) for the purpose of studying the effects of the 

 use of green manures and fertilizers on the productivity of neglected lands which 

 have been cropped in hay continuously for a long period of years. 



The yields of hay on the plats seeded in 1921 were obtained for the seasons 

 1922, 1923 and 1924. The yields of hay on the plats seeded in 1922 were 

 obtained for the seasons 1923 and 1924. The data have been tabulated, but 

 because of the variations in the different plats the probable errors were too 



