THE STRUCTURE OF A FEATHER 267 



wood, and strips of card cut out to the shape of the 

 barbules may be fixed to saw-cuts made in the sides of 

 the bars. A little trouble bestowed upon the details of 

 the model will not be thrown away ; it all helps 

 the perfect understanding of a beautiful and intricate 

 mechanism. In the decorative plumes of many birds 

 the barbules are undeveloped or lose their hooks, and 

 the barbs then become free. A piece of a peacock's 

 feather mounted as a lantern-slide, makes the arrange- 

 ment of the barbs and barbules quite plain. Double- 

 shafted feathers are easily shown to a number of 

 persons, if mounted in the same way. 



A feather cannot be mastered until its development 

 has been studied. Something may be seen of the 

 development of a feather by examination of a moult- 

 ing Bird, and it is seldom that a bird is not moulting 

 some of its feathers. The new feathers may be seen 

 pushing up through the skin, each enclosed in a thin 

 outer quill, which crumbles gradually away from the 

 tip downwards, and allows the barbs to expand. But 

 the easiest way to get developing feathers is to ex- 

 amine unhatched chicks. Chicks removed from the 

 egg after incubation for nine days and upwards, 

 provide excellent material. But few of my readers, I 

 fear, can command a supply of developing chicks, or 

 know how to investigate them. The work is mainly 

 done by thin sections through the artificially hardened 

 tissues. 



There are two layers in the skin of Vertebrate 

 animals, an outer layer (epidermis), which is cellular 

 and neither vascular nor sensitive ; and an inner layer 

 (dermis], which is abundantly supplied with vessels 



