A source of possible error in the method of deduction from these results must be taken into consideration. The observations at lightships and lighthouses are mostly made when untoward circumstances bring the birds within range of vision, and on dark and foggy nights cause them to strike the light in great numbers. What is their normal course when no great migration wave or "rush" is observed? Are the few passing stragglers noted all that go by this route in fair weather? The same uncertainty must be applied to the observation of passing birds in inland localities. The immense numbers which do pass is shown by the observation of large movements, when as occasionally happens some check to normal migration leads an army of birds to a dangerously low altitude, or when high winds hold up a portion of the host on our coasts; but even these multitudes must be small compared with the millions of birds which annually pass from zone to zone unseen. The few or many birds we meet with, either on the coast or inland, resting on passage, may represent a lost or wandering party of stragglers or weaklings from a vast army which has passed over; they may or may not be on the route or course normally followed by the majority. The cartography of bird migration is a study in itself. Mr Abel Chapman, describing his experiences in the Mediterranean, says--"For forty hours we were passing across (or beneath) the lines of an army of migrants--say 500 miles in width; yet not a sign did we see, save only the wreckage--the feeble that fell out by the way." On April 10th a sudden bitter northerly gale sprang up, and two hours later the steamer was the goal of hundreds of birds, no longer able to face the adverse wind. These were blue-headed wagtails, swallows, martins, pipits, wheatears, nightjars, and lesser kestrels. He thinks that the strong ones may have passed on but all the others perished (12). * * * * * CHAPTER IV THE HEIGHT AND SPEED OF MIGRATION FLIGHT In the last chapter reference was made to the great height at which birds may fly on migration. Certain species, even comparatively weak-winged ones, appear normally to fly high, whilst others, often birds with pointed wings and great aerial powers, usually proceed at low elevations; but there is still much conjecture as to the actual altitude reached by any migrants.