98 RECLAIMING LOST NITROGEN. 
with a good yield. Upon examining the roots of the plants they are 
found to have developed tubercles, while the control plants, watered 
with sterilized pure water, do not develop tubercles. These facts 
of course indicate that in the soil infusion some agencies are 
present which stimulate the development of tubercles and the 
consequent fixation of nitrogen, and that the power of absorbing 
atmospheric nitrogen enables the plant to recover from the nitro- 
gen-hunger stage. 
The Tubercle Bacteria. These facts naturally suggest that 
bacteria, or other microorganisms, are the cause of the tubercles. 
Microscopic study of the tubercles shows a 
somewhat perplexing structure. The tubercle is 
the result of the excessive growth of the cells of 
the root, but they are filled with peculiar bodies. 
During the early growth of the tubercle, long, 
thread-like sacs appear, which force their way 
through the cells (Fig. 25). These filaments seem 
to be hollow tubes which contain smaller bodies, 
pouches in the tuber- somewhat like bacteria. As the legume in- 
represents ^wo^ells creases in size these bacteria-like bodies undergo 
with the sacs pene- a transformation in shape, growing larger and 
tratingthemOS/e/aw). 
branching somewhat, so as to form structures 
like those shown in Fig. 26. These are called bacterioids, and they 
are characteristic of the tubercles of legumes. The next step was, 
naturally, to isolate these bodies and study them by bacteriological 
methods. It is easy to isolate from the tubercles bacteria that will 
grow in culture media, and these organisms were named B. radicicola. 
Experiments with the bacteria thus isolated have been extensive 
and, on the whole, satisfactory, though occasionally they have been 
conflicting. It has been proved many times that tubercles can be 
produced upon legumes by the cultures thus obtained. Legumes 
have been grown in sterilized soil and watered with bacterial in- 
fusion from these cultures: the usual result has been the growth of 
abundant tubercles and the fixation of nitrogen. Some striking 
experiments have been made with germinating peas. Such peas, if 
kept moist and warm, will grow for several days, sending out their 
