190 BOTANICAL GAZETTE [MARCH 
of a previous epoch, there may be seen low dunes built by Arcto- 
staphylos, Juniperus, and Potentilla fruticosa, small dune excres- 
cences on a general dune substratum. The life-history of these 
dunes appears to have been as follows. First a stage of dune 
formation by such plants as Salix, then, as vegetative propaga- 
tion allowed these plants to encroach more and more lakeward — 
upon the broad beach, new dunes were formed nearer and nearer 
the lake. Or, perhaps, there was a recession of the lake anda 
consequent extension of the beach into new territory; these 
new dunes may have thus been formed farther and farther lake- 
ward, keeping pace with the advancing shore line, In any event, 
the interposition of a new row of dunes between the lake and 
those first formed essentially changed the life conditions on the 
latter. The row of dunes nearest the lake serves as a windbreak. 
The first row catches most of the drifting sand, and the second 
catches most of what remains. This fact makes it possible for 
the slow dune-formers to inhabit the more inland of the ridges 
Expressed in other words, the high-grade or primary embryonic | 
dunes encroach upon the lake, not by the actual advance of al 
individual dune in that direction, but by the formation of new 
ones; in like manner the low-grade or secondary embryonic 
dunes encroach upon the former type, using the topographic 
form of the primary dune as a base of operations, and building 
new dunes as low excrescences upon the old. 
The advance of a dune area toward the lake, as just described, 
shows how the coastal belt of dunes may grow wider as the yeals 
pass by. In another place it will be shown how they may also 
Stow wider by the actual advance of an individual dune 7 
the land. In concluding the section on the embryonic dunes # 
may be well to emphasize in another way the radical differene® 
between the two types that have been discussed. Theit intimal” 
gradation, as shown at Beaver island, is by no means the unive 
fact. Perhaps it is even more common for one of the primary 
type to leave its original habitat and wander across the oa 
as an active dune. The secondary type never has that histor” 
an Arctostaphylos dune almost always grows into a heath. 
