18 BULLETIN 934, U. S. DEPARTMENT OF AGRICULTURE. 
vanced that damping-off may be a valuable selective agent in nursery- 
grown stock for forest planting, eliminating the weaker individuals 
and thus insuring the vigor of the trees which go into the forest 
plantation. This is a possibility which must be considered. It is by 
no means certain* however, that escape from damping-off is correlated 
with permanently superior vigor. It is believed that temperature, 
moisture, and other environmental factors, which as yet are very im- 
perfectly analyzed, together with the age of the seedlings and the 
presence or absence of virulent strains of the parasites, are much 
more important factors than inherent differences in individual re- 
sistance in determining whether or not seedlings are destroyed. Evi- 
dence from inoculation experiments and from field observation, sup- 
ported by data of the sort presented in Table I, indicate that damp- 
ing-off ordinarily does a considerable part of its damage by killing 
the sprouting seed before emergence from the soil, while the graphs 
in figure 9 show that of the loss which occurs after emergence in un- 
treated beds a large part occurs very early in the life of the seedlings. 
Observation of the clean sweeps which the disease commonly makes 
in the immediate neighborhood of infection foci (figs. 3, 4, and 5) 
indicates that either before or just after the seedlings break through 
the soil none of them have any considerable resistance to the really 
virulent strains of the parasites, which are believed to be the ones 
responsible for the major share of the damping-off. 
Even if there should be found to be an appreciable selective value 
in damping-off, this would not be a valid argument against control 
by seed-bed disinfection for the following reason: The graphs show- 
ing the course of damping-off in treated plats in figure 9. together 
with the decided differences in germination between treated and un- 
treated plats, indicate that the very early damping-off is more com- 
pletely controlled by disinfectants than the later damping-off. This 
early damping-off which the treatments so largely prevent is the 
part of the loss which has the least possibility of selective value. 
The later damping-off is rarely controlled at all thoroughly by disin- 
fectants. As shown by the graphs, it is often even heavier on the 
treated than on the untreated soil. It is the part of the loss which 
is most likely to have selective effect. At this stage beds are not 
taken clean, as earlier; only seedlings which are below normal re- 
sistance succumb. The damping-off in disinfected beds seems there- 
fore at least as likely to have true selective value as that which 
occurs in untreated beds. 
The only way in which the effect of damping-off as a selective 
agent can be positively determined will be to compare through sev- 
eral subsequent years the growth rate, or survival after transplant- 
ing, of trees from beds which suffer seriously from damping-off 
with the growth of trees from the same lot of seed in seed beds in 
