84 BULLETIN 934, U. S. DEPARTMENT OF AGRICULTURE. 
phytes had been added and those which received no saprophytes 
showed a slight increase, proportionally as well as in the , absolute 
figures. At the end of the 36 days the survivals on all the pots were 
counted separately. The average number of seedlings per pot were 
as follows : 
Without Pythium: 
Without saprophytes, 42.8±2.3 ; with saprophytes, 52.1±1.1 
With Pythium : 
Without saprophytes, 30.1±2.4 ; with saprophytes, 48.1±1.4. 
The difference in the survivals in favor of the pots with sapro- 
phytes in the first case is 9.3 ±2.5, three and two-thirds times its 
probable error. In the second case it is 18.0 ±2.8, more than six times 
its probable error. 
In general, it appears that in this experiment the inoculation of ster- 
ilized soil with saprophytes gave the seedlings some protection both 
against clamping-off due to accidental infection with unidentified 
parasites and from the additional loss caused by light inoculation 
with Pythium deharyanmn. The indication is, as would be expected, 
that only part of the favoring influence of heat sterilization of soil 
on introduced P. deharyanum is immediately due to the elimination 
of competition with other fungi. If a mixture of different bacteria 
and fungi had been added to each of the pots instead of but one or 
two organisms to each 5-pot unit, the effect might have been more 
marked. 
It will be noted that for all the groups (fig. 18), whether with or 
without Pythium inoculation and with or without added parasites, 
the frequency polygon is asymmetrical, indicating by its shape, as 
did the frequency polygon of survivals in pots inoculated with Rheo- 
sporangium (fig. 17), that with infections which do not kill all of 
the seedlings the selection tends to be by pots rather than by seed- 
lings. In other words, in pots in which the parasites succeed in kill- 
ing any of the seedlings, they usually kill a considerable number. 
The tendency is illustrated not only by inspection of the graphs, but 
by the variability of the different groups. The greater variability 
in survivals between different pots was in both cases in the groups in 
which both the damping-off after emergence and the survival per- 
centages indicated the largest loss. The percentages of seedlings 
damped-off during the entire 36 days following emergence and the 
coefficients of 'variability of the survivals of the individual pots at 
the end of that time are as follows : 
Without Pythium: 
Without saprophytes, 15.5 per cent damped-off; coefficient of variability, 
39±4.2 per cent. 
With saprophytes, 11.1 per cent damped-off; coefficient of variability. 
28±1.6 per cent. 
