DEVELOPMENT OF PLANTS 



171 



gliding movement and this is supposed to be due to the expan- 

 sion and contraction of minute strands of protoplasm that pro- 

 ject through the pores of the valves (Fig. 96, C, p). 



(a) Reproduction of the Diatoms. — It will naturally be asked 

 how can these plants living in glass houses, grow? As the valves 

 become changed to silica naturally any increase in size must 

 cease. Nevertheless the cells reproduce with great rapidity and 

 in a very interesting manner. Through the growth of the living 

 substance the valves are pushed apart and the cell contents di- 

 vides, forming two diatoms with but one valve each. A new 

 valve that fits into the old valve is soon developed on the un- 

 covered side of each diatom and two complete diatoms are thus 

 formed (Fig. 98) which may become free at once or remain at- 



IQ3ltt 



Fig. 98. Diagram illustrating the divisions of a diatom and the resulting 

 reduction in size. The brackets connect the two diatoms that were formed 

 by each successive division. — After Pfitzer. 



tached and by further division form the odd groupings or colo- 

 nies shown in Fig. 94. Since the valves are of unequal size, i. e., 

 one fitting into the other, the two diatoms formed by the division 

 must vary in size and it will be seen that the majority of the off- 

 spring will become greatly reduced in size as a result of the 

 rapid and repeated divisions. That this reduction may not go 

 on too far, when a certain minimum size has been reached the 

 cell contents, with or without dividing, throws off the valves 

 entirely and grows to the full size of the diatom, when new 

 valves are formed and the diatom is ready to repeat the dividing 

 process. In some cases we find quite a different method of re- 

 production. Two diatoms become enclosed in a jelly-like mass, 

 into which the cell contents, with or without dividing, is dis- 

 charged. These naked cells now unite in pairs and the bodies 

 thus formed finally grow into diatoms (Fig. 99). This latter 



