THOMAS KERR 39 



the major cell axis. The two opposite orientations within the same wall 

 frequently show optical compensation and the parts of the wall adjoining 

 the intercellular spaces may appear isotropic. When one examines a series 

 of cells grading from parenchyma to coUenchyma, the second orientation 

 parallel to the cell axis gradually becomes the dominant one. This 

 parallel orientation is the arrangement of the cellulose in the thickened 

 corners of coUenchyma. Thus two types of cellulose orientation, one in 

 which the fibrils are perpendicular to the major cell axis, and a second 

 in which the fibrils are parallel to the major cell axis, may be seen in cells 

 that are undergoing rapid enlargement. In neither case is there any 

 appreciable change in the orientation of the microfibrils during growth. 

 Therefore it follows that growth in any direction is independent of the 

 major orientation of the microfibrils within the wall. 



The changes that occur in the primary wall during growth have 

 recently been considered by Frey-Wyssling (5). He postulates that the 

 microfibrils of cellulose form a network which is held together by lateral 

 forces, presumably by hydrogen bonding or some equivalent force. 

 During growth the lateral forces holding the fibrils must be loosened, 

 the empty spaces enlarged, and new strands must be interwoven. Frey- 

 Wyssling's hypothesis will explain the plastic extension or growth of the 

 wall. On the other hand, it is extremely difficult considering our present 

 knowledge to see how a wall built in this manner could possess the elastic 

 extensibility of primary membranes. Primary walls are known to expand 

 and contract as much as 30 per cent with turgor changes in the protoplast. 

 Such elastic properties could be explained if the wall were composed of an 

 elastic or rubber-like substance or it could also be explained if the wall 

 were composed of a highly hydrophyllic material that would change 

 in volume and in surface area with the water relations of the protoplast. 

 Cellulose is known to possess rigid molecules having only very limited 

 elasticity and does not display anything of a rubbery nature. Further- 

 more sheets of cellulose do not change appreciably either in volume or 

 surface area in different neutral solutions with a range of four or five 

 atmospheres. If the wall were composed of a skeletal framework of 

 cellulose, it would be expected to be a nonelastic structure similar to a 

 sheet of cellophane. 



The picture is unfortunately more complicated. Properties of cellulose 

 have been evaluated from dried material, and it is now recognized that 

 natural cellulose, that has never been dried, possesses somewhat different 



