90 BOTANY 



sion into two and the gradual separation of the two from 

 one another. Here we see the simplest form of repro- 

 duction following growth, each of the two protoplasts 

 resulting from the division possessing the same proper- 

 ties as the original cell. When in some cases the 

 processes of growth and division became very rapid, a 

 second division might well take place in the two new 

 cells before they had become separated. There would 

 arise in this way by degrees a new form of plant, one in 

 which the processes of division continued, but the cells 

 did not separate at all, so that a chain or filament of 

 cells resulted. They would keep their individuality at 

 first at any rate, being almost as independent of each 

 other as if they had separated, their attachment being 

 mainly mechanical. We find a very large number of 

 water-weeds, or algae, existing in our ponds and streams 

 to-day which have just this structure. 



The maintenance of this thread-like structure depends 

 on the cells always dividing in such a way that the new 

 cell walls arise across the length of the thread. It seems 

 certain that this course was soon departed from by some 

 of the plants and that some divisions arose at right 

 angles to the others. This led to the formation of a flat 

 plate of cells, still only one cell thick. A plant with this 

 structure would meet with very little greater difficulty 

 than the thread-like forms ; the requirements of the 

 cells would be the same, though the plant might easily 

 grow very much larger, the number of its protoplasts or 

 cells often amounting to many hundreds. Still the cells 

 would be all alike and probably all independent, for 

 water would have access to each and no further pro- 

 vision would be called for. Plants of this type of 

 structure are still common among marine seaweeds. 



A great complexity, however, would arise if the 

 protoplasts began to divide in three planes, each at 

 right angles to the other two. No doubt this was not 



