STATISTICS OF METASPERMAE. 617 
It will be necessary to observe one or two points in this 
division. In the first place it must be recognised that not all 
of the families in any of these groups are of equivalent distri- 
bution. In Group A, for example, have been included such 
families as are represented in both tropical and extratropical 
regions of both eastern and western hemispheres. A family 
of which the range answered to such a description might nev- 
ertheless be very much more limited in its distribution than 
one which might be found in almost every continent or island 
—as, for illustration, the Juncacee. The groups are therefore 
intended to be and are somewhat elastic. Again, it is some- 
times thought advisable to include the same family in two, or 
even three groups, in order to give the proper notion of its 
range. For example, the Sarraceniacee includes three genera, 
Sarracenia, Chrysamphora and Heliamphora. The first two are 
limited to North America, one being Atlantic, the other Pacific. 
The third is found in British Guiana. Under these conditions 
of North American preponderance it seems proper to enter the 
Sarraceniacee as North American. But since a genus is devel- 
oped in South America it seems proper, too, to enrol the fam- 
ily among the Western Hemisphere forms. Third, it will be 
noticed that Cosmopolitan families belong also to the next tive 
divisions; Extratropical families include also the Northern 
extratropical. North American families are included in the 
Western Hemisphere group. Evidently, then, the general 
intent of the classification into groups is to give not total 
range, but distinctive range. Wesee, then, how the Juglandacew 
may be classed as Northern Extratropical, while the Saxifraga- 
cee, being represented also in the southern hemisphere, are 
more properly placed under the wide group of Extratropical 
families. 
The following table will indicate the distinctive range of 
Minnesota valley families: 
