1883.] 221 [Jeffries. 



three, while the cells divided off from them as a rule have only 

 one nucleolus. At no time during the development of the scales 

 have I found the distinction between the cells to be wanting. 



About the end of the twelfth day marked changes have oc- 

 curred in the scales both as regards form and structure. They 

 no longer form ridges pointing up from the surface, but folds 

 towards the tip of the limb. Kerbert states that by this time 

 they have begun to grow in a bilateral symmetrical manner, 

 while the down feathers continue to grow in a radial symmetrical 

 manner. The cells (fig. 15 b) next to the epitrichial layer have by 

 this time begun in various spots to assume a fusiform shape, as 

 seen in section, their nuclei at the same time becoming elongated. 

 The number of cells between the epitrichial and mucous layer 

 is very variable, there being most on the tarsus and least on the 

 toes. All, however, are as yet protoplasmic and have none of the 

 structural characteristics of horn. The mucous layer itself is 

 much the same as before, the cells varying from columnar to 

 cuboidal, according as they have just divided or are about to 

 divide. The epitrichial cells are spread out and come much 

 nearer a true cell-coat than on the tenth day. 



By the fifteenth day the flap of the scale has become quite 

 long, and marked differences between its outer and inner surface 

 also exist. The epitrichial layer has become distinctly a layer of 

 cells ; though the cells are still very variable in size and shape. 

 Beneath the epitrichial layer is the "granular layer" of Kerbert 

 and others. This he figures as one layer of polygonal cells, with 

 large nuclei and marked granular contents. According to my 

 observations, which cover all the distinct parts of the epiderm 

 except the cloaca, this layer is only the cells divided off from the 

 mucous layer, just after the epitrichial layer, that do not become 

 converted into horn-cells. 



In thickness this layer varies much, in some parts it is one cell 

 thick, in others many cells thick, and in others does not exist 

 at all. 



Below this granular layer is the horn layer of very different 

 thickness in different parts. It is thickest about the middle of 

 a scale and thins out on the inner side. At this time the horn 

 layer varies from one to four cells deep where it exists. The 

 cells appear fusiform in section, have thick walls and small cavities 



