14    Wisconsin Academy of Sciences, Arts and Letters 
information came from those who had conducted winter feed- 
ing operations. Some farmers had partridges in their fields 
without knowing it; others knew of the first outposts miles away. 
Many a day's questioning yielded only discarded data; again an 
hour's discussion yielded a clear picture of half a county. All 
data were accumulated on county maps. 
During the summer of 1937 John Beule, one of my students, 
mapped the spread in Dodge County and adjoining parts of 
Fond du Lac and Sheboygan. His data are incorporated in Fig. 1. 
Both my own work and Beule's support the hypothesis that 
the Wisconsin partridge "front" advances by three mechanisms: 
(1) By slow yearly overflow into adjacent unoccupied ter- 
ritory. This is the "ripple" type of spread. 
(2) By salients suddenly thrust out into unoccupied territory 
and then slowly amalgamated with the main front. 
(3) By isolated outposts of population thrown far ahead 
of the main front. These may enlarge and eventually 
coalesce with the main front. 
While the long-time trend is one of aggression into new 
territory, this trend is the net resultant of many local retreats 
and halts as well as advances. Salients or outposts are thrust 
out only to encounter bad seasons or adverse range and die. 
"Bubbles" or vacant spots are left behind the main advance, and

may not become populated for a decade (for example, most of 
Ozaukee County). Some are submarginal range and persist in- 
definitely (for example, Horicon Marsh and parts of the kettle 
moraine). Others of marginal quality doubtless disappear and 
reform with varying population pressure. But despite these 
local defeats, the partridge front has, during the 30 years since 
their introduction, advanced steadily across the fertile farm- 
lands eastward to the barrier of Lake Michigan, southward to a 
juncture with the Illinois populations, westward into the prairies 
of the driftless area, northwestward to the border of the sands. 
and northward into the rich clays of the Green Bay region. 
What will ultimately halt their advance no man can yet say. 
The finer details of spread pattern during the early years are 
lost. The outward movement can be depicted only in terms of 
crude 10-year contours, such as those for 1920 and 1930 (Fig. 1).