THE ACQUISITION OF ATMOSPHERIC NITROGEN. 117 



the outside of the roots of various legumes. Some of the organ- 

 isms succeed in forcing their way into the tissues of the young 

 roots, though they are not able to pierce the older roots. For a 

 while they may remain in the root as free bacteria, but the plant 

 plasma seems to exert an injurious influence upon them, for 

 very soon a thin membrane is formed around the bacteria 

 masses, inclosing them like a pouch. Prazmowski thinks that 

 this membrane is a product of the bacteria themselves, formed 

 for the purpose of protecting them from the injurious action 

 of the plant tissue. The bacteria which do not succeed in get- 

 ting into one of these pouches soon cease to grow and degener- 

 ate into irregular forms like the bacteroids which appear later in 

 greater numbers. The bulk of the bacteria, however, become 

 enclosed in the membrane, after which they continue their 

 growth with much vigor. The pouches begin to grow into 

 threadlike masses, and these make their way among the cells of 

 the root. The thread branches more or less as it lengthens and 

 its various filaments grow through and between the cells, soon 

 permeating the root with a fine, branching filament, which looks 

 much like the mycelium of a mold. It was this bacteria pouch 

 which was first seen by Erickson, and which previous observers 

 regarded as the hypha of some low fungus. Instead of being 

 a mycelium growth of a mold the thread is nothing more than a 

 large, branching colony of bacteria inclosed in a thin membrane. 



"The growth of this colony of bacteria among the cells of the 

 root stimulates these cells to an unusual growth. They multiply 

 more rapidly than usual, and thus soon produce a swelling on 

 the root which is the beginning of the tubercle. While this 

 rapid multiplication of root cells is going on, the bacteria pouch 

 continues to grow, and swells out into rounded vesicles within 

 the cells which lie at the center of the forming tubercle until 

 most of them become filled with these expanded portions of the 

 bacteria thread. Meantime the root cells of the plant have been 

 rapidly growing, and form around the cells containing the bac- 

 teria several layers of smaller cells, which develop into a hard, 

 corky covering forming a coat around the tubercle. This seems 

 to be impervious to the bacteria thread, and confines the bac- 

 teria within its limits. 



"The bacteria colony now undergoes a change. Although 

 Prazmowski has not been able to follow the details of the pro- 



