THE METHOD OF HOMING PIGEONS. 765 



in preparation for the following experiments for nearly four 

 months. 



On May 29 th six pigeons were caught at random and carried 

 in a cage, the top and sides of which were made of open-mesh 

 wire netting to allow free vision, to the top of a hill about half a 

 mile southeast of " home." The birds were liberated singly, and 

 the course flown by each in returning to loft may be seen by a 

 glance at Fig. 1. At the same time, six other birds, also taken 

 without selection, were carried to the same place in a basket 

 closely wrapped in a heavy black shawl. Tracings of their home- 

 ward flights are reproduced in Fig. 2. 



By comparing the two figures may be seen the influence of 

 vision upon directness of return flight. In Fig. 1, five of the birds 

 are seen to start toward home at once ; one, a young bird out of 

 the loft for the first time, flies in the wrong direction a short dis- 

 tance, turns sharply about, and alights upon the first house on the 

 line toward home. Three pigeons fly home without preliminary 

 circling. Fig. 2 shows not a single direct course. Two begin 

 circling in the wrong direction. One of these persists in his false 

 bent to the extent of searching over the whole city of Madison. 

 In the case of Fig. 1, the birds, most of them, see home very soon 

 and fly directly to it. Fig. 2 shows more of the method which a 

 pigeon adopts in covering unknown territory. 



By comparing successive tracings of the same bird, may ' e 

 seen the effect of education i. e., improvement in good birds, 

 failure to improve in poor ones. The first is well shown in Fig. 

 3, the three different lines representing three successive flights of 

 pigeon ISTo. 2. As this is a typical case, let us follow its develop- 

 ment a moment. No. 2 has been carried from the loft in a covered 

 basket to the hill half a mile distant. Not having any idea as to 

 which direction he has been taken, he naturally starts out the 

 wrong way. His first thought, if I may so speak, is to look around 

 for some familiar object. So at a he turns about and is occupied 

 with looking over the landscape, while he describes a number of 

 semicircles back and forth around the place of starting. He evi- 

 dently sees nothing until his widening circles bring him over a 

 group of buildings around a farmhouse at h. Among these 

 is a red barn, and close to the loft is a red house. This looks 

 quite like home, and he swoops down toward it. On closer ex- 

 amination, however, he fails to discover the familiar barn, the 

 pigeonholes, and alighting board. After making four circles 

 low down over and among these buildings, he rises high in the 

 air again to take another look. All to the south and west are lake 

 and marsh and woods, with but few farmhouses visible. Toward 

 the north he sees a cluster of houses South Madison. He looks 

 them over in the same way. From c he catches sight of another 



