There is wide-spread misconception concerning the speed at which birds normally fly, and even regarding the speed they can attain when occasion demands, as when closely pursued by an enemy. It is not unusual to hear accounts of birds flying "a mile a minute." While undoubtedly some birds can and do attain a speed even greater than this, such cases are exceptional, and it is safe to say that even when pressed, few can develop an air speed of 60 miles an hour. They do, however, have two speeds, one being the normal rate for everyday purposes and also for migration, and an accelerated speed for escape or pursuit; this in some cases may be nearly double the normal rate of movement. Nevertheless, the effort required for the high speeds could not be long sustained, certainly not for the long-distance migratory journeys that are regularly made by most birds. [Illustration: B4749M Figure 5.--Migration of the cliff swallow; a day migrant that, instead of flying across the Caribbean Sea as does the blackpoll warbler (fig. 11), follows around the coast of Central America, where food is readily obtained.] The theory that migrating birds attain high speeds received encouragement from the German ornithologist Gätke, who for many years, made observations on birds at the island of Heligoland and published thereon in 1891 (_21_). He postulated that the blue-throat (_Cyanosylvia suecica_), a species of thrush smaller than the American hermit thrush, would leave African winter quarters at dusk and reach Heligoland at dawn, which would mean a sustained speed of 200 miles an hour; and that the American golden plover (_Pluvialis dominion_) flew from the coast of Labrador to Brazil in 15 hours, or at the tremendous speed of 250 miles per hour. These conclusions are now considered unwarranted by most ornithologists. Sportsmen also often greatly overestimate the speed at which ducks and geese fly and sometimes attempt to substantiate their estimates by mathematical calculations, based upon the known velocity of a charge of shot, the estimated distance, and the estimated "lead" that was necessary to hit the bird. If all three elements of the equation were known with certainty, the speed of the bird could be determined with a fair degree of accuracy. The majority of the ducks that are reported as killed at 40, 50, or even 60 yards, however, actually are shot at distances much less than estimated. To sight along a gun barrel and estimate correctly the distance of a moving object against the sky is so nearly impossible for the average gunner as to make such calculations of little value.