ROOT-NODULES OF LEGUMINOSAE. 105 
Beyerinck found swarmers in minute nodules which were still inclosed in the mother root. He 
divides the root-nodule organisms into groups and varieties as follows: 
Group I.—This contains the larger more hyaline colonies. Growth absent or difficult on meat 
peptone gelatin. Growth is favored by cane-sugar or grape-sugar. Swarmers are very minute. The 
bacteroids are two-armed, globose, or pear-shaped. Meristem is always present in the nodules. ‘The 
primary bark of the nodule is closed. Slime threads are distinct. The following forms belong here: 
B. radicicola, vars. fabae, vicia-hirsutae, trifoliorum, pisi, lathyri. 
Group II.—Colonies more cloudy white. Growth better on meat peptone gelatin. Swarmers 
more rod-shaped, somewhat longer. Bacteroids like the bacteria, that is, seldom branched. Slime 
threads absent or little developed. Mostly no meristem in the nodules (Robinia an exception). 
Three types occur: (1) Phaseolus type; (2) B. radicicola, var. lupini; (3) Robinia type. 
In Vicia faba, as the bacteroids are exhausted the color of the cytoplasm changes from reddish 
to intense green. The bacilli from this plant when grown in Faba stem gelatin in a cool place (cellar) 
were alive and motile at the end of a year. Active cultures can be obtained from all parts of the 
nodules which have been exhausted by the bacteria. They are present in a living condition therein 
in great numbers. The result is 
quite different when the host 
empties out the contents of the 
bacteroids. Then it is more and 
more difficult to get any bacterial 
growth from the meristem. The 
longer the bacteria remain in the 
nodules the more bacteroids occur. 
Beyerinck found saprophytes 
in the nodule tissues mixed in 
with B. radicicola and named at 
least two—B. luteo albus and B. 
agglomerans. Another green 
fluorescent form thought certainly 
to come from the nodule was 
identified as B. fluorescens putidus. 
A form resembling B. radicicola 
and found in certain nodules was 
first named B. radicicola lique- 
faciens, but subsequently Beyer- 
inck came to regard this as an 
intruder having nothing to do 
with their formation. This lique- 
fying organism was afterwards 
called Bacterium beyerinckii by 
Trevisan. 
The bacteroids are found in 
other parts of the roots than the 
nodule, but less well developed, 
e. g., in the root-hairs and epi- 
dermis cells. Beyerinck never 
found them in parts above ground, except once in a stem of Vicia faba where inoculated by hypo- 
dermic injection. 
The bacteroids are always derived from the bacteria. They occur in old cultures as well as 
in the nodules. The swarmers easily pass through the walls of the Chamberland filter. 
When fresh nodules are put into water at room temperature this water clouds first with a mixture 
of bacteria, of which B. radicicola is the chief. Later, when the nodules decay, other bacteria appear. 
The tissues of legumes have a strong attraction for this organism, as is shown by the fact that 
in such roots placed in the water any little cracks or wounds are immediately occupied by this 
organism and the intercellular spaces flooded with it. These roots may be considered as a bacterial 
trap apparatus. 
The infection of the living pericambium of the root must take place through pores, possibly 
*Fic. 35.—Poured-agar plates of Bact. leguminosarum from bean, introduced to show effect of repeated freezings: 
a, Contents of a loop before freezing—several hundred colonies per sauare centimeter; 5, Contents of a similar loop 
of culture fluid after 7 freezings—less than one colony per each 2 square centimeters. Each freezing lasted half an 
hour; time between freezing short, 7. ¢., only long enough to thaw out tube in cool water and make necessary plates. 
Round colonies are on surface; spindle-shaped ones are buried. 
Fig. 35.* 
