SOILS FOR SEED PKODUCTIOX. 147 



the French types just mentioned. In most cases they 

 have a capacity of only a few tons, the laboratory selec- 

 tion taking place in the spring. 



These silos are sunk ij feet in the earth and are 

 four feet wide, the necks slanting upward; about 

 18 inches of earth are piled on top; the covering 

 should be flat, so that some moisture from rains, etc., 

 may readily penetrate. Knauer claims that if the root 

 be moistened it will keep better in silos; precaution 

 alone being necessary to prevent stagnant water. Dur- 

 ing very dry seasons, the piles of mothers may be 

 watered; the water carries the earth down to surround 

 each beet. When the total covering of silos with 2\ 

 feet of earth is finished, during very cold winters, it 

 may be found desirable to still further cover with barn- 

 yard manure. The opening of the silos depends upon 

 the method of selection. Knauer says that selected 

 beets in well conducted silos should be placed one 

 against the other, and not one on top of the other. This 

 covering, upon general principles, would seem to be a 

 bad practice; for the weight of earth has a tendency 

 to crush the beets and thus bring about considerable 

 changes in the entire pile; this is the reason why many 

 advocate straw and a thin covering of earth. 



Considerable experience is needed to know just 

 when to place the beets in silos. Better select a very dry 

 day; if rainy, the beets siloed in their wet condition 

 would in most cases undergo fermentation. On the 

 other hand, if exposed to the sun too long,, the roots 

 wilt and the chemical selection that follows would be 

 very misleading, as the sugar percentage would appear 

 to be higher than the reality. Under all circumstances, 

 loss of sugar percentage occurs during the months the 

 roots are kept; hence the reason why the chemical 

 selection should take place early. Arguments in favor 

 of late selection, showing which roots have keeping 

 qualities, are not as reliable as one would wish. Expe- 



