81 
of the several trees must be made in a particular way 
in order to give the most reliable result, but this is a 
point of detail into which [ will not now enter. 
Now let me show you what results the investigation 
brought to light. 
IL. The very considerable fluctuations which appear in 
the yearly growth of the oak-wood of the region under 
investigation must in great part certainly be due to meteo- 
rological influences. 
Proof. The proof of this proposition is furnished by 
the figures I now throw on the screen (figs. II, III, IV, 
V). The tree-growth has been graphically represented 
in this figure. Each vertical line represents a determined 
year. On this line (see number at the top of the plate) 
has been plotted the growth in the year 1700 on the 
next line that of the year 1701 and so on. The growth 
itself was expressed as a fraction of the average growth 
and as average growth was considered the average 
growth of the surrounding 15 years. !) The number of 
trees on which the results depend have been marked on 
the curves. Owing to the different ages of the trees, 
this number generally decreases as we pass to earlier dates. 
Now then: what proves that the fluctuations in the 
growth-curves are not due to local causes (depending, 
say, on cutting down of surrounding trees; on artificial 
irrigation questions etc.) is their parallelism. Generally 
l) That is the mean growth of the 7 preceding years, the year 
itself and the of following years. 
In Figs. I— V the average growth has been represented by 20. The 
interval between two consecutive lines thus represents 5 percent of the 
average growth. Thus in Fig. Il the growth in the year 1834 is 31 
percent above the average. In Figs. VI and VII the average growth is 
taken — 10 and the interval between two horizontal lines represents 
consequently 10 percent of the average growth. 
Recueil des trav. bot. Néerl. Vol. XI. 1914. 6 
