THE DIVISION OF Till': PROTOPLAST 73 



at 19°, the prophase occupies 36 to 45 minutes, the metaphase 7 to 10, 

 the anaphase 15 to 20, and the telophase 20 to 35; total, not inchidin^ 

 interphase, 78 to 110 minutes. In the brown alga, Sphacelaria, growing 

 at nearly the same temperature, the process requires less than half as 

 much time. Mesenchyme cells of the chick growing in tissue cultin-es 

 at 39°C. pass through prophase in 5 to 50 minutes, usually more than 30; 

 metaphase, 1 to 15, usually 2 to 10; anaphase, 1 to 5, usually 2 to 3; 

 telophase and cytokinesis, 32 to 133; total, 70 to 180 minutes. Choroidal 

 cells from chick embryos and cartilage cells from adult fowls carry through 

 their division in about half this time under like conditions. In fibroblasts 

 from a l-da}- mouse in a 2-da3' tissue culture, about 10 minutes elapse 

 between the initiation of the equatorial furrow during anaphase and the 

 completion of cytokinesis. In the development of the male gametophyte 

 and gametes from microspores of the water fern Marsilea, there are nine 

 successive cell divisions and then a transformation of certain cells into 

 spermatozoids. All this has been observed to occur in as short a time 

 as 10 to 12 hours; hence the divisions and the intervals between them 

 must be of short duration. 



The Shape of Cells in Tissues. — Obviously the shape of a tissue cell 

 must be related to its internal differentiation and to the more general 

 conditions pervading the tissue or organ of which it is a part. It is 

 nevertheless a matter of considerable interest to determine what shapes 

 tissue cells assume when conditions are as simple and uniform as 

 possible. 



Investigations in this field have shown that in a flat epithelium or 

 epidermis the cells in surface view tend strongly to be hexagonal in outline. 

 If a flat Plasmodium with its nuclei scattered at random in one layer 

 were divided into uninucleate cells by walls with minimal surface area, a 

 hexagonal pattern would result. Such a pattern is seen in the cucumber 

 epidermis. When one of the hexagonal cells divides, the new wall forms 

 stable three-rayed intersections with two opposite walls of the hexagon, 

 the daughter cells being pentagons, and two of the adjacent cells becoming 

 heptagons (Fig. 50). With subsequent divisions, chiefly of the larger 

 cells with more than six sides, the number of sides per cell varies still 

 further in the tissue, but the general average remains not far from six. 



A similar play of forces at cytokinesis in a uniform three-dimensional 

 Plasmodium would result in the formation of a mass of cells each having 

 14: sides and trihedral intersections, tetrahedral angles being unstable 

 and rare. If the space w^ere uniformly divided into polyhedral cells 

 with equal volume, minimal surface area, edges of equal length, and no 

 intercellular spaces, each cell would have the form of an orthic tetrakaide- 

 cahedron, which has 8 hexagonal and (> (luadrilateral faces. Cells in 

 uniform tissue mas.ses such as pitii approach this 14-sided form. Further 



