In order to give a survey of the proportional numbers in which the 
two observers agree and in which they differ, the cases in which 
there is agreement within the limits of the probable errors, are 
expressed in the subjoined field of squares. 
Cases in which the proportional number — does not differ in 
9 
Z and Hs, within the limits of the probable errors. 
A definite order in these agreements is not to be perceived as yet. 
We have tried to discover whether a greater agreement might 
perhaps be obtained, if instead of the proportion of the olfaction* 
values those of the logarithms of the olfaction-values were taken. 
The subjoined tables account for this. 
If in these tables the values are looked up that in the tables to 
the right and left are about the same* it soon appears that no manner 
of regularity can be observed. 
Hereupon Mr. Hermanides has tried a vectorial representation of 
