ba [ ASSEMBLY, No. 33.] 
or the variation can be claimed as no greater in germination than in 
vegetation or vice versa, and hence we are enabled to logically assume 
that in onions we must anticipate extreme variations between seeds 
from the same package as possibly occurring dependent upon the draw- 
ing of the samples. 
We may further assume logically that our figures, as given in our 
Tables J and II, are not gathered from a sufficient number of seeds 
used to admit of other than most general conclusions. 
