204- BULLETIN OF WISCONSIN NATURAL HISTORY SOCIETY. VOL. 2. NO. 4. 
The topography to the north of Madison shows that Lake 
Mendota at one time had an outlet northwest into the Wisconsin 
River, and that a httle to the northeast of this area there is a broad 
valley that once connected Rock River with Green Bay (45). A 
possible conclusion suggests itself now as an explanation of the 
two directions noted at Madison, namely, that they are the result 
of these two physiographic features. 
These conclusions point to a feature of special interest which 
offers a most important field for investigation. In general the 
families Mniotiltidse and Turidse, and the order Limicolae breed 
in the far north. Among the others certain species breed abund- 
antly up to a definite latitude, beyond which the number of species 
and individuals of these more southern birds decreases directly 
as the distance increases. Granting Weismann's theory in so far 
as migration is a development, resulting from dispersion, as we 
must, and Merrian's theory that there are two centers of disper- 
sion in North America (46), it would seem evident from the fore- 
going considerations that birds dispersing; from the north south- 
ward, in the same region, may have been influenced by different 
topographic features from those that dispersed from the south 
northw^ard. 
From a detailed study of the geographic distribution of birds, 
their lines of migration, and the degree of perfection which the 
phenomenon has attained in the different groups, we may arrive 
at data which will lead to the correct conception of the origin of 
migration, and from which we may even be enabled to work out 
the evolution of the phenomenon in the different groups. 
Prof. Baird has traced the lines of dispersion of a number of 
birds of eastern Xorth America westw^ard and northward, almost 
to the Pacific Coast in the northern part of the United States. 
He found that in dispersing they followed the Missouri and 
Platte Rivers from the Mississippi River and, what is most re- 
markable, that they followed these same rivers in migrating from 
these summer homes to their winter homes in Mexico, instead of 
moving directly southward through California (47). This is ex- 
cellent evidence that the routes followed in migration are the lines 
of dispersion followed when birds first mov^d out from their 
original faunal area. Moreover, the main highways at present 
are situated along courses that are especially favorable as lines 
45. Chambeiiin. Geol. of Wis. Survey of 1873-1879, vol. 1. p. 284. 
46. Merriam. U. S. Dept. of Agri., North Amer. Fauna Bull., No. 3, 
pp. 19-22. 
47. Baird. Am. Jour, of Sc., 1866, p. 78. 
