200 BULLETIN OF WISCONSIN NATLTIAL HISTORY SOCIETY-. TOLT 2. XO. 4. 
troit and Minneapolis ; from Lake Forest there were two ; and the 
writer, with the assistance of Mr. Wegemann, obtained three at 
Beloit, while Dr. Libb^ conducted the observations at Madison. 
The data received from ^Michigan deserve special mention for the 
care with which they were collected. 
From the telescopic observations it appears that birds are quite, 
we might even say, exceptionally dependent upon their respective 
migratory routes. Curves drawn from such data representing the 
numbers of birds flying in periods of fifteen minutes show that 
•physiographic features are followed very closely as guide-lines 
(Plate I.). The Beloit curves show a sudden increase in the num- 
bers of birds observed as the telescope gradually comes to point 
more nearly over the course of the Rock River. The observatory 
at Beloit is situated on a slight elevation just east of the river, so 
that the telescope points away from the river in the early evening. 
The river at this point flows very nearly south, while just outside 
of the southern border of the cky it takes a course a little west of 
south. These curves again show the extreme importance of a 
water course in migration. In the foregoing consideration it has 
been stated that a great many of the birds of Wisconsin enter the 
state along the course of the Rock River. The difference between 
the number of birds flying over the immediate course of the river 
and those flying to one side of it, as shown by the plates, indicates 
that topographical lines are followed very closely here. There is 
an exceptional concentration of the number of birds flying over 
the immediate course of the river. 
While the curves drawn for ]\Iadison do not bring out this 
feature in as striking a contrast as do those for Beloit. the obser- 
vations show that there is a decrease in the number flying when 
the field of observation has passed beyond the ^Madison Lakes and 
and Yahara River System. By reducing the different plots for 
Madison to the same number of hours after moon-rise, we obtain 
a comparative view of the numbers of birds flving over the same 
physiographic features (Plate IL). Although the curves show 
considerable diversity, they seem to reach maximum points be- 
tween two and three, four and five, and seven and eight hours after 
moon-rise. This, however, is not universally true. The lake re- 
gion about ^Madison, with its extremely diverse topographical feat- 
ures, makes the problem a most complex one in this region. From 
investigations carried on here for over four years it is quite saf^ 
to conclude that the greatest flights are found to occur when the 
telescope points over the lake system — that is, before the moon 
' has traveled over the skv a distance of five hours. Plate IIL, 
