NO. 2 DIMENSIONS OF FLYING ANIMALS — GREENEWALT 3 



Note also that soaring birds, the albatross particularly, are not ex- 

 traordinary in relative wing area, falling generally in line with the 

 small passerines. 



WING LENGTH AND WING AREA 



Figure 5 shows the relationship for birds, figure 6 for insects. The 

 birds fall into a very consistent pattern, but here the differences for 

 soaring birds become more apparent. The albatross, for example, has 

 a very long wing per unit area, as does the frigatebird and booby. This 

 means simply that for soaring birds the wings are long and narrow, a 

 condition essential for good aerodynamic stability, which does not re- 

 quire per se a large wing area. 



In figure 6, the insects show their unusually large "scatter." We have 

 models ranging from the long, narrow wing of the fruitflies and crane- 

 flies to the broad stubby wings of the butterflies. The proportionality 

 constant in the equation relating wing area with the square of the wing 

 length varies through a factor of 5. For birds the variation is scarcely 

 a factor of 2. 



Figure 7 shows data for bats. One sees that these data are very self- 

 consistent and that the constant of proportionality is quite close to that 

 for birds. The flying model is similar, much more so than the appear- 

 ance of the two classes of animals would lead one to expect. 



WING SPREAD AND WING LENGTH 



In virtually all ornithological handbooks the wing length as given is 

 not the length of the whole wing, but that of what is called the "hand," 

 viz, the distance from the wing tip to the first articulated joint. This 

 practice arises out of the great difficulty in measuring total wing length 

 or wing spread from bird skins, as compared with the relative ease of 

 measuring the length of the "hand." Figure 8 shows Magnan's data 

 on wing spread plotted against the measurements of the length of the 

 "hand." It is essential here to use data from a single investigation since 

 precise measurement of wing spread is greatly influenced by the tech- 

 nique of the particular observer. We see that the two hands average 

 62 percent of the wing spread. The "scatter" is not great, a tribute to 

 Magnan's self-consistency. 



WING AREA AND WING WEIGHT 



In dimensional theory, the weight of the wing should be proportional 

 to the cube of its length, or to the L5 power of its area. Figure 9 shows 

 the relationship for insects and birds. We see that wing weight is pro- 

 portional not to the 1.5 power, but to the 1.67 power of the wing area. 

 Since we have previously shown wing area proportional to the square 



