Fa 
others. [I hoped to do this in a more extensive and 
thorough way than it was done in my lectures and to add 
some materials about the period 1879 up to the present 
time. — Other work, however, has always made me 
delay this plan, so that finally I felt that nothing would 
ever come of the matter unless [| resolved to publish just 
my lecture in Pasadena. 
About the lecture itself Ï wish to make a few remarks. 
Ist In expressing numerically the relation between 
wood-growth and rain, the most scientific way would have 
been to compute correlation-coefficients. Meanwhile I have 
on several occasions found that the real significance of 
this coefficient is very obscure even to biologists who are 
accustomed to express their results in this way. The 
cause of this lies undoubtedly in the fact that the coefficient 
had not been correctly defined. I have tried to remedy 
this defect myself (see Monthly Notices of the Royal 
Astron. Society, April 1912), but [I have reason to think 
that the simple definition there given is not yet very 
widely known. This being the case I have preferred not 
to change anything in the way of expression adopted in 
the lecture and to retain it in the Notes. This way may 
be less scientific but it is more generally understood. 
2nd. For the greater part of the lecture use has been 
made only of part of the wood collected. For the regions 
summarized in figs I—V all the trees measured were 
included. But I have not discussed the wood collected 
near Bonn and Wesel, in Holland, in Oldenburg and 
near the Baltic !)} Of what was obtained from the 
neighbourhood of the Ems only one batch, that which 
shows the 12 year-period in such a remarkable way, was 
considered. The trees from these regions which were 
1) An exception has been made for the investigation of the influence 
of temperature. See Note II. 
