( 301 ) 



The differences between a and b, and between c and </ give the 

 effecl of the substitution ui' number of summerdays for mean-tempe- 

 ratures, the differences between u and b, and between c and </ the 

 effect of a shift of 15 days. It appears that the improvement in the 

 concordance of t tie signs is largest for the transition from h to </: 

 in the transition from c to '/ in t lie first place the improvement also 

 in the quantitative agreement of' the deviations of mortality and 

 number of summerdays is remarkable. Finally the agreement in case 

 (I may be called so satisfactory, that lint little doubt remains whether 

 the high temperatures must be considered as the cause of the increased 

 mortality of infants. 



Hence the negative result arrived at by Prof. Sai.tkt and Mv. Fw, ken- 

 berg was due partly to the use of mean temperatures instead of 

 temperature frequencies, partly to their non considering a retardation 

 of the mortality with respect to the cause of death. 



We will now test the agreement in case (/ also for Utrecht and 

 Zealand. 



UTRECHT (de Bilt). 



ZEALAND (Flushing) 



