1913] Holmes: Ectodcrmic Epitlieliiim of Tadpoles 161 



increases, although small yolk granules may be seen in the cells 

 of the free swimming tadpoles. Parallel changes in the yolk 

 content of cells occurred in pieces of tadpole tails kept several 

 weeks in lymph. Many j)reparations in which the ectoderm cells 

 were crowded with yolk globules when first put up had assimi- 

 lated most of the yolk and became nearly tran.sparent after 

 being kept for two or three weeks in vitro. Some pieces, as will 

 be more fully described in another paper, underwent a consider- 

 able increase in volume. The ectoderm had extended over the 

 cut surface so as to form a perfectly continuous investment and 

 the cells had become thinned out and widened to two or three 

 times their original diameter, while the original yolk content 

 had been almost entirely used up, leaving the cells nearly trans- 

 parent. 



With the disappearance of the yolk the ectodei-mal cells are 

 rendered capable of spreading out more widely. Those that 

 become free usually remain more or less rounded in outline 

 unless they come into contact with a solid surface. The older 

 ectoderm cells then manifest an extraordinary capacity for 

 spreading. The area in contact with the substratum becomes 

 increased by the emission of short fine and very transparent 

 pseudopods. By this means the cells are enabled to extend in 

 all directions at the same time. The margins of these cells are 

 iLsually irregular, and one may usually observe changes in their 

 outlines and follow various stages in the spreading of the cyto- 

 plasm. Cells frequently spread out on the cover slip so as to 

 become of extreme thinness. Except for a small granular area 

 around the nucleus the cytoplasm may be so very transparent 

 that it is invisible under ordinary conditions, and the true extent 

 of the cell can be ascertained only by observing its delicate amoe- 

 boid boundary. Neighboring cells as a consequence of this spread- 

 ing frequently come into contact, and where there are several 

 of them near together they may form a continuous membrane. 

 In plate 7, figure 5, is shown a small part of a secondary mem- 

 brane formed in this way. The cells composing it had originally 

 become scattered through the plasma ; then they began to flatten 

 out, the process continuing until no open spaces remained between 

 the cells. In this case the seeondarv membrane so formed 



