INSECT MIGRATION 233
dragonflies conform to an aerial route along the north slopes of the seaward dunes, or in the valley, with a precision so exact that one may take his stand beforehand and find them passing directly overhead or even within reach of one's hand.
Upon broken, irregular coasts quite diflferent behavior is seen : some insects follow one path, others another. So confused and peculiar do some of these flight- ways become, in fact, that they have led to some erroneous conclusions. Just such an instance seems revealed by a letter written in 1901 by Mr. Henrv Bird, a resident of Eve on the north shore of Long Island Sound. In correspondence with a Canadian entomologist he says :
I used to find time for a day's fishing now and then and in a row-boat would anchor a half-mile out from shore on some submerged reef. . . . And one is sure to Jye struck with the number of butterflies constantly coming from the Island [Long Island] to the main shore. We might expect the supply to be largest on the mainland, and that emigration would be going the other way, yet it is invariably as first stated. By far the largest number seen will be Archippiu ("monarchs"), next Papilios (swallowtails), and Colias philodice (clouded sulphur). "Monarchs^' seem on the wing continually during the last half of the season, and their flight over the Sound waters seems most methodical. But in this case — they are going north, mind — ^they fly singly.
The Canadian scientist, Mr. J. A. MoflPat, then concludes the insects were gathering in the north preparatory to a southward migration.
So curious a happening seemed worthy of further investigation, par- ticularly as insect migrations along the Sound shores had only been predicted by the writer and never definitely proved. So Milford Beach, several miles farther east than Rye, was chosen as the place of observation.
How curious was the phenomenon that presented itself almost im- mediately upon my arrival during a late August afternoon! For not only were "monarchs," swallowtail butterflies and dragonflies flying from over the water and toward the shore — ^that is, toward the north- west — ^but swallows, as well, were accompanying them. All alike con- tinued inland and disappeared from sight beyond the houses and woods in that landward vicinity. So here, almost at once, was a repetition of the peculiar spectacle reported from the Rye shore fourteen years before. What is the solution?
Former Long Island studies, where swallow and insect routes coin- cide, gave the key to the mystery. For, upon looking far out over the water, it was seen that the swallows were traveling more nearly parallel to the shore, and seemed to be winging their way from some outjutting point of land farther east (rather than from Long Island), in an attempt to pass over the waters of the cove or bay. But on being driven southward bv the brisk northwest wind thev, as well as the insects, which had evidently set sail from a similar point, were forced to turn
�� �
