STATISTICS OF METASPERMIC PLANTS OF THE 
MINNESOTA VALLEY. 
Value of statistics. In the following pages such numerical 
data as have seemed most necessary for an understanding of 
the distributional characters and physiognomic features of the 
Minnesota valley Metaspermae have been compiled. It is pro- 
per first of all to call attention to what has well been stated by 
De Candolle—that somewhat too great an air of mathematical 
exactness is conveyed even to professional botanists, by sta- 
tistics of such a nature. The apparent rigidity of the calcula- 
tions inspires a mistaken notion of rigidity in the distribution, 
On the contrary, however, the statistics are not truly indicative 
of such mathematical exactness of distribution as their appear- 
ance would imply; they are toa certain extent of the nature 
of averages or estimates. There are in their compilation, too, 
certain sources of error; for example, in the table which gives 
the north, south, east and west preponderance of species in the 
North American continent, the entry of any species is a matter 
of judgment from published geographical lists, and either the 
data of the lists may be incomplete or erroneous, or the judg- 
ment may be faulty. In order to correct such error it has 
been my effort to compile the statistics from several points of 
view. Under such method it is presumed that they will mutu- 
all correct each other and the general results will thus come to 
be of definite reliability. 
Again, there is the danger that the personal equation of 
error, indicated in mistakes of counting or calculation, will per- 
mit a degree of vitiation that were better avoided. Although 
each group of statistics has been carefully compiled and the 
results checked in such a way as to guard the general results 
to what extent may be possible, it will readily be seen that a 
single small error would propagate itself unlimitedly, if by 
accident it should creep into the preliminary count. 
