83 
felled near Treves could be carried through. The figure 
thrown on the screen shows the result. 
In the last figure (Tab. IV, Fig. IX) the full line represents 
the tree-growth in percents over or below the average 
growth. Each interval represents 5 percent of the mean 
growth. 
The dotted line similarly represents the height of the 
rainfall in spring and summer over or below the average. 
Only the fluctuations have here been reduced about two 
times !). 
We see that the parallelism is quite close. 
In fact I find that on an average, if at Treves the height 
of the rain during spring and summer has been 100 per- 
cent over the average, the growth of the trees here in- 
vestigated has been 48 percent over the average. The 
result has been extended somewhat for years in which 
no rain measurements have been made, because we pos- 
sess some further countings of the number of rainy days. The 
result has been shown in Tab. IV, Fig. VIII. Computation 
shows that on the average, when the number of rainy 
days in spring and summer is 100 percent above the mean 
number, the wood-growth is 40,6 percent above the mean. 
From a similar comparison made between winter-rain and 
wood growth, we find hardly any trace of an influence. 
Ï know that this result is in contradiction with what 
so high an authority in forest matters as Ebermayer 
maintains, who judges that winter-rain and snow is of far 
greater importance for our forests *). But Ï cannot see 
how, for the Treves wood under consideration, we can, with 
any possibility, escape from our conclusion. Meanwhile we 
must take care not to extend our conclusions farther than 
1) See note Il. 
?) See for instance Ebermayer, Die physikalischen Einwirkungen 
des Waldes etc. Berlin 1873, p. 197 en 198. 
