12 
measured are about equal in number to those discussed 
in the present paper, but they are mostly much younger. 
They show relatively little regularity. I suspect that most 
of this wood grew not in extensive forests, and that it is 
mainly in such forests, well situated in respect to subsoil- 
water supply that we must look for uniformity. 
3d. Though the nature of the soil and the situation of 
the trees with respect to rivers etc. must be of greatest 
importance, my data about these points are very scanty. 
The cause is that most of my specimens were obtained 
from wood-merchants, who could not of course be 
expected to take particular interest in these matters. 
Even in a few cases there remained a doubt of a year 
about the date at which the trees have been cut down. 
Happily these cases are rare. By careful comparison 
with trees from the same neighbourhood about the date 
of which there is no doubt, they could usually be assig- 
ned without trouble to their proper epoch. 
What informations I obtained (about the trees discussed 
in this paper) is embodied in the following summary: 
(See pag. 73.) 
4th, The results for the different trees were combined 
into normals in such a way that no tree would have a 
predominant influence. In order to reach this aim smaller 
weight was given to those trees of which the breadth 
of the ring is very considerable. Where the average 
growth of the same tree is markedly different for diffe- 
rent periods of their life, these weights were correspon- 
dingly altered. Where for some larger period the breadth 
of the ring is exceedingly small, no good measures could 
be made at all. Together with the somewhat different year 
of cutting this accounts for the fact that sometimes the 
number of trees on which our normals rest is somewhat 
smaller in the later years than in the former. The num- 
bers of trees which have contributed to our normals have 
