Great differences of opinion have been expressed about the speed of migrating birds, and here again Gätke's estimates, on account of the weakness of his arguments and his immense presumption, cannot be seriously considered. There are but few measured speeds, and most of these, except perhaps the ducks and geese referred to already, are of birds travelling at low elevations. Many birds, especially day-migrating swallows, hooded crows and other birds, frequently travel at slight elevations; it is not unusual to see birds at sea flying a few feet only above the waves. Mr W. Eagle Clarke, whose systematic observations demand the profoundest respect, again and again urges that the direction of the wind has little effect upon migration, but that the force of the wind may make migration impossible. At the Eddystone, where he spent a month in the autumn of 1901, he noticed birds passing at heights varying from 20 to 200 feet, all flying southwards. He concluded that "the wind is certainly the main factor in migration meteorology--I am convinced that the _direction_ of the wind is, in itself, of no moment to the emigrants, for they flitted across the Channel southwards with winds from all quarters" (16). When the velocity of the wind, however, was above 28 miles per hour (a fresh breeze, force 5 on the Beaufort scale), no migration was observed. Allowing this, and also taking into consideration the undoubted fact that birds are frequently held up by strong winds on the shore before starting oversea journeys, is there any proof that they do not actually avail themselves of fairly steady strong winds when they reach the upper air?