202 BULLETIN OF WISCONSIN NATURAL HISTORY SOCIETY. VOL. 2. NO. 4. 
per cent that did not take a general northeasterly direction. The 
course of the Detroit River is southwesterly, and Mr. H. S. War- 
ren writes that it seems evident that the birds were following the 
river. The regularity of the direction maintained should dispel all ' 
doubts to the contrary. An interesting feature now presents it- 
self in connection with the peculiar topography just northeast of 
Detroit. The river here gradually broadens to form Lake St. 
Clair. While 66 2-3 per cent of the birds maintained a northeast 
direction, 21 2-3 diverged slightly to either side in about equal 
proportion, from which we might infer that the latter were begin- 
ning to direct their courses along opposite shore lines of Lake St. 
Clair. If this be the case, it shows how very closely birds follow 
topographic lines. Then, again, it is rightly maintained that birds 
can see great distances, from which we should expect that unless 
thev follow the coast lines of the lake very closely the telescopic 
observations collected here would show a great diversiy in the 
directions of flight, since the birds would then cross from Lake 
Erie to Lake St. Clair rather promiscuously. Under such circum- 
stances there would be very little evidence of concentration along 
special topographical lines. 
A comparison of the plots for Beloit and Madison show that 
at Beloit the maximum in the curve is reached much more sud- 
denly than at Madison (Plates L to III.). This, again, accord- 
ing to our theory, is readily explainable in each case by the dif- 
ference in the surroundig topographv. The Lake System at Mad- 
ison is broad and has the topographic features that influence lines 
of migration extending over a wide area. The Rock River at Be- 
loit is narrow and the surrounding topography rather uniform. 
The curves for the latter place consequently rise more suddenly 
and at the same time vastly higher, the migrating stream being 
confined to narrower limits. 
The birds noted to either side of the more important physio- 
graphic features are probably such as have departed from the 
main highway and are following courses of lesser importance, in 
many instances such as will probably lead them to a summer home 
in the vicinity. In fact. Plates IV. and V., illustrating the direc- 
tion of flight at Beloit, seem to show that outside of the main 
highway (i. e., during the early part of the evening, when the 
telescope pointed away from the river) there is greater diversity 
in the directions followed. When the telescope points away from 
the main highwav, the fact that fewer birds were seen and that 
the directions followed were more diverse would point to the con- 
