~ 
162 BOTANICAL GAZETTE [MARCH 
after. Pioneer formations usually are hydrophytic or xerophytic, 
mostly xerophytic in arid climates, and more equally divided in 
moist climates. For example, the last retreat of the glacial ice 
left in our northern states a vast tract made up essentially of hills 
and hollows, the hollows, if deep enough, with lakes. The first 
vegetation of the hills was xerophytic, and the first vegetation of 
the hollows, hydrophytic. Finally, except on the higher hills and 
in the deeper hollows, these first plant formations gave way step 
_ by step to the climatic tundra, and, as the climate became amelio- 
rated, this in turn gave way to climatic coniferous forests, and then 
to climatic deciduous forests as they exist today. So far have the 
higher hills and the deeper hollows lagged behind the less extreme 
habitats in their development that there are still to be found many 
places which continue to have pioneer formations, though, of course, 
they differ greatly from the original pioneer formations of the 
tundra. 
While the general trend of vegetation-is from diversity toward 
_ uniformity, it must not be supposed that complete similitude is 
ever reached even under like climatic conditions. There are species, 
for example, in the culminating forest of New England which do not 
occur in Ohio, and species in Ohio which do not occur in Illinois; 
southward the difference is even more pronounced. And yet it 
cannot be denied that from the Maritime Provinces to Minnesota 
and south to the coastal plain the ultimate forest in its larger 
features is of a single type; the percentages and even the kinds of 
dominating trees may differ, but the aspect is essentially the same. 
Much more diverse from one another than are the beginnings OF” 
the ends are the intervening stages. Our northern lakes, for 
example, differ much less from one another in their floristic compo- 
sition than do the swamps to which they give rise. The initial 
stages of a rock upland in Tennessee and in northern Michigan ar¢ 
much alike, both in aspect and in floristic composition; the te 
minal stages in these two widely separated districts are even more 
alike, but the intermediate stages are very different, northern Michi- 
gan having nothing at all comparable to the oak stages in the ves" 
tational development of eastern Tennessee, and the latter rego? 
being without the complex coniferous stages of northern Michigan. 
