1 6 DESCRIPTIVE BOTANY. PART I. 



into straight lines, the circles will become hexagons 

 (fig. 4.). If a number of spheres, of equal size, be in 

 contact, each may be touched by twelve others 

 (fig. 5. a ) ; and if 

 the whole be subjected 

 to pressure, so that 

 their surfaces may be- 

 come flattened at these 

 twelve points, the 

 spheres will become 

 rhomboidal-dodecahedrons (fig. 5. ft). But, as the vesi- 

 cles which compose the cellular tissue are never exactly of 

 the same dimensions, the polygonal forms which they 

 assume will not be so strictly regular as the geometric 

 figure we have just mentioned. Still, there is often a 

 very marked approximation towards such a regularity ; 

 more especially in those parts of the plant which are the 

 best developed, or have been most securely defended, 

 as in the case of the pith, from the influence of disturb- 

 ing causes. Where the vesicles are elongated, the dode- 

 cahedrons assume the character of rectangular prisms, 

 terminated by four-sided pyramids, whose faces replace 

 the angles of the pyramids at various degrees of inclin- 

 ation to the axis (fig. 6'.). If sections be made through 

 these, by planes paral- 

 lel and perpendicular to //^ 

 the faces of the prisms, 

 they will exhibit either 

 hexagonal or quadran- 

 gular surfaces, accord- 

 ing to circumstances, as 

 a simple inspection of the diagrams will be sufficient to 

 show. Cells of these forms may be so aggregated 

 (fig. 7.) as to fill space as completely as the hexagonal 

 prisms of the honeycomb; but as the extreme regularity 

 here delineated is never actually attained in nature, the 

 cellular tissue becomes every where penetrated by small 

 cavities, by which an intercellular communication is 

 maintained throughout the mass. These channels are 



