54 CELL-DIVISION AND GROWTH II. i 



sliding or rotation, or both. Even two cells will glide on one 

 another, as two soap-bubbles will not. In complexes two three- 

 surface lines may unite to form one four-surface line, a behaviour 

 the very opposite of that exhibited by soap-bubbles. 



It appears, then, that in the living egg of the Frog (and other 

 radial and bilateral types) there are factors which overcom- 

 pensate, to use Roux's expression, the purely physical factors by 

 which the behaviour of the oil-drops is governed. These organic 

 factors are that division is slow, and begins on the outside ; that 

 the direction of division determined by the yolk is persistent ; 

 that the cell contents are neither perfectly fluid nor perfectly 

 structureless ; that the cells being different, their surface tensions 

 may be of different magnitudes, and the whole system, therefore, 

 not homogeneous ; and that the cells possess a more or less solid 

 rind or membrane, the rind which becomes wrinkled transversely 

 to the furrow when the cell divides. 



It would seem that this rind is an important factor, for if 

 Roux's experiments be repeated with drops of albumen suspended 

 between xylol and oil of cloves, to which a little alcohol has been 

 added, it will be seen that each drop gets a superficial membrane, 

 and that by these membranes adjacent drops adhere. In fact, 

 such drops behave more like the cells of the egg than do the oil- 

 drops. Thus, a small cell goes towards the inside, or the outside, 

 according to the way in which the division is made, and, after 

 a horizontal division of four equal cells, the upper remain super- 

 imposed upon the lower. 



At the same time, it is apparently because the cells have this 

 surface film, which the oil-drops have not, that they are able to 

 flatten against one another as soap-bubbles do ; while, on the 

 other hand, it is because the film is solid that the cells are 

 unable to move upon one another and adopt the geometrical 

 arrangement seen in systems of soap-bubbles. 



There is still another kind of cell-movement to which brief 

 reference must here be made, since it is found in one type of 

 segmentation at least. In the segmenting eggs of some Platy- 

 helmia (Triclads), Ascidians (Salps), Echinoderms (Asteroids), and 

 Coelenterates (Oceania), the blastomeres have been seen to com- 

 pletely separate from one another, afterwards reuniting. Roux 



