��INSECT MIGRATION 229
"'^^d from the northern shore of western Lake Erie. For they pred- ^to<i some great, and probably specific, source of supply in tl^ o ^^^an region. As a matter of fact, the territory immediately to ^ ^orth was first considered the probable source of this living butter- ^ K ^^^G&m. But few, if any, records could be found in that region. 'A^^t: previous trail then could these myriads have followed other than ^^ * ^fDrthem shores of the Lakes, as these were perfectly natural lanes ^ ^vel if the behavior already observed in the east could be duplicated ^ For it would seem that, in this region too, most of the butter- ^^^ coming down from northeastern Canada would be held back and Reflected by these water barriers, just as they are in Connecticut and Long Island by the Sound and ocean shore.
Only the evidence then was needed. This was readily supplied bj reprts of resting swarms at Port Stanley, Toronto, Montreal (for the route may even extend along the St. Lawrence Valley), and by the interesting description of "thousands, even hundreds of thousands of the red-winged butterflies resting on the shores of Lake Ontario at West Point," which was communicated to the waiter by Miss Esther Steele, of Pittsford, N. Y. Moreover, this same observer reports that the butter- flies never appear in any great numbers in this home region, almost directly opposite West Point, but near the southern shore, indicating that the insects rarely, if ever, cross the lake. This supposition is even further confirmed bv another observer who did see a swarm at Cleveland, on the southern shore of Lake Erie; but who supplements his record with the statement, " they are seldom seen in this locality." Evidently, then, the "monarchs" usually avoid these great water barriers, pre- ferring to traverse the 500 mile route already outlined, only to flood southward in the released stream which finds its outlet along Point Pelee and the contiguous islands leading to the Ohio shore, thence the flights probably continue directly southward, as shown by the record at Circleville; but after that point is reached our knowledge of their further course is lost.
Still another great route has been indicated along the west shore of I^ke Michigan by reason of the swarms reported at Racine and Chicago. This flight-line, also, probably represents a released stream whose source lies still further north in some undetermined region hevond the Great Lakes. Its outlet toward the south, however, is shown by numerous reports throughout central and eastern Illinois, and, prob- ably, by the flight still farther down country at Alton, where this tenuous thread, linking the insects of the temperate and the tropic zones, is lost.
But the wide highways of the Great Plains and the West Central States offer the most frequent reports of remarkable butterfly spectacles. In fact, they are so common in Minnesota, Iowa, Kansas, Oklahoma