The Cellular Basis of Growth 49 



3. The early wall formed by the cell plate, and certainly the phragmo- 

 some which precedes it in vacuolate cells, are not at first continuous films 

 and would thus not follow the law of least surfaces. 



4. In many cases, as often in the unequal division that cuts off a 

 stomatal mother cell, the new wall is at first straight instead of curved 

 and becomes curved only later, as the turgor of the cell increases. 



5. Frequently, as in growing cork layers, the new division wall is laid 

 down exactly opposite a partition wall in an adjacent cell so that four 

 walls do come together at a point (p. 195). This also happens in tissue 

 which is to form aerenchyma and in which the cells are in regular rows 

 with cross walls opposite. Here, however, at the point where the four 

 walls meet, a small air space (which later may enlarge) is commonly 

 formed by the pulling apart of the walls so that the wall angles do tend 

 to reach the theoretical 120°. 



6. In a system of film bubbles increasing in number by the formation 

 of new walls, the equilibrium least-surface configuration is reached by a 

 shifting of the wall positions within the film system. This involves some 

 gliding or sliding of the bubbles in relation to each other. Such a change 

 could happen in animal tissues where the cells are free to move about, at 

 least to some degree, but would be impossible in most plant tissues, where 

 they are cemented to one another. 



For these reasons it is clear that the theory of surface forces alone is by 

 no means sufficient to explain all the facts as to the position of new cell 

 walls and the planes of cell division. Other physical factors are doubtless 

 involved in determining these events. Among them pressure is important. 

 Kny ( 1902 ) found that pressure applied to a dividing cell forced the 

 mitotic figure into a position in which its long axis was oriented at right 

 angles to the direction of the pressure, and the new wall consequently 

 was parallel to this direction. This fact, incidentally, makes an important 

 contribution to our knowledge of the character of the cytoplasm, at least 

 at this time in the history of the cell. If the cytoplasm were essentially 

 fluid, pressure from without should not change the orientation of struc- 

 tures in it but would do so if the cytoplasm had a structural framework. 

 Other evidence for the conclusion that walls are formed parallel to pres- 

 sure on the cell can be found in the cortex of the young stems of many 

 woody plants. Here the cells tend to be elongated tangentially, presum- 

 ably because of the pressure exerted by the expansion of the vascular 

 cylinder below through cambial activity. If these cells divide again, 

 radial walls, parallel to the direction of cambium pressure, are often to be 

 seen. As Kny points out, however, in cambial cells, which are presumably 

 under radial pressure, division is chiefly periclinal (at right angles to the 

 pressure) instead of anticlinal. This he attributes to "inner factors." In the 



