80 
was cut down to know the year corresponding to all 
the rings. 
For my investigation I began by collecting discs of 
trees from various parts of Germany and Holland. In this 
lecture Ï will mainly confine myself to the forests along 
the Main, the Moselle and to some of the forests not 
far from the Rhine between Worms and Bonn. On 
each cross section under examination Î first drew a 
pencil line beginning at the centre and crossing all the 
rings at right angles. Along this line | measured care- 
fully the breadth of all the rings and they were recorded 
against the years in which they grew. In order to 
improve the accuracy I usually repeated the same thing 
along several other radii. 
We thus obtain at once a fairly good measure of the 
yearly growth of the tree during the whole of its life. 
Having done the same for all the trees in the region 
now under consideration, [| found at once that there was 
a considerable agreement in the growth in contiguous 
forests. This is not always the case, proving that what 
the trees register depends on their situation. It is pretty 
evident that this must be so. For a tree standing onthe 
border of a lake at a pretty constant level must not be so 
strongly dependent on the quantity of rain as other trees, 
not so well situated, and so in other cases. But for the 
forests now under consideration the parallelism is very 
striking and other and more direct data being not avail- 
able, I took this as a proof that the trees in the different 
forests must have been growing unter much the same 
conditions. This enabled me to combine all my results 
into four definitive results, (see Tab. IT figs. IT, IT, IV, VI) 
three of which represent approximately the tree-growth 
for an area about equal to !/; part of Holland, while 
the fourth is restricted to a smaller area. The combinatfbn 
