STATISTICS OF METASPERMAE. 751 
necessity of explaining such a fact by the tension-lines and the 
law of ejections, and reciprocally the fact of distribution is of 
value as evidence of the soundness of the law. 
The marsh-plants exceed the total flora in northernness 
easternness and westernness, but fall behind in southernness. 
Not yet fully distributed as are the aquatics, they indicate bet- 
ter the influence of the continental tension. Marsh plants in a 
given region of the northern hemisphere may be expected to 
present distal rather than central characters, for as has been 
discussed above, the tendency to adopt the morassic habitat is 
a distal or tension-line phenomenon. As might be expected 
where different forces are acting to determine the percentages 
one will often partly neutralise another. Itis seen, for exam- 
ple, in the table,that the northern and southern percentages of 
the metachlamydeous marsh-plants are very close together, 
while the eastern and western percentages are not so close. 
This is just the reverse of the condition among the Metachla- 
mydeae as a group, in the Minnesota-valley flora, and indicates 
the selective influence of habitat upon range. In the drier-soil 
element, on the other hand, the condition of the total flora re- 
appears and is accentuated. 
The drier-soil plants lead the total flora only in southernness, 
while in northernness, easternness and westernness they fall 
behind. Of the element, the Archichlamydeae lead in northern- 
ness, the Metachlamydeae in southernness, the Monocotyle- 
dones in easternness and the Metachlamydeae in westernness. 
The Archichlamydeae of the drier-soil are less northern, more 
southern, less eastern and less western than in the total ele- 
ment. The Metachlamydeae of the drier-soil are less northern 
more southern, less eastern and less western than in the total 
element. The Monocotyledones do not differ from the two 
groups mentioned, in this particular. That both the eastern- 
ness and westernness of the drier-soil element should be de- 
creased in all taxonomic groups indicates the wide east and 
west distribution of the two elements the removal of which 
leaves it as the residuum. 
In general the study of the tables which indicate the range 
of the physiognomic elements will add weight to the belief that 
the three taxonomic groups are of different and distinct mean- 
ing in the distribution. Space scarcely permits as exhaustive 
an analysis as might be useful but enough has been noted in pass- 
ing to show how a further and more complete analysis should 
properly proceed. Careful examination of the tables and com- 
