124 BACTERIA IN RELATION TO PLANT DISEASES. 
The surest method of seed-inoculation is first to swell the seeds and then inoculate them. The 
soaking of the seeds must, however, never take place under water but in moist sand. 
The name bacteroid, it is said, originated with Brunchorst (1885) who implied thereby that these 
. peculiar bodies while they resemble bacteria in size, shape, and staining properties have really nothing 
to do with bacteria. Beyerinck, by cultivating them on artificial media, showed their true bacterial 
nature. Prazmowski was the first to produce nodules by inoculation with pure cultures (1890). 
Frank then abandoned his earlier views and described a micrococcus which he had isolated from 
the nodules and which he named Rhizobium leguminosarum. ‘This he claimed differed in form and 
behavior on gelatin from Beyerinck’s organism, and was the true cause of the nodule formation. He 
claimed that these bacteria were taken up by the bacteroids which he considered albuminous bodies, 
and released only when the albumen was absorbed by the plant. He termed the mixture of bacteria 
and plasma Mykoplasma. Afterwards, Frank admitted that the bacteroids developed from the 
nodule bacteria but claimed that they were only involution forms serving as storehouses for the 
albumen which the plant absorbed. ‘This view has prevailed until recently. Hiltner maintains that 
this is a false idea. They are not involution forms. Frank does not explain whence bacteroids obtain 
their albumen when plants are grown in nitrogen-free sand. Further, according to Hiltner, the 
increased growth of the plant takes place before any absorption of bacteroids occurs and the amount 
of nitrogen fixed by the plant during a season may exceed by 100 times that contained in the nodules. 
Hartleb, it is said, succeeded in producing bacteroids in artificial media, and attributed the 
alteration to the action of phosphoric acid. Hiltner does not agree with this latter view though he 
states that there is no doubt that bacteroids may be produced by artificial means, as he himself 
obtained them with root-extracts from legumes. 
Hiltner has repeatedly opposed the idea that the bacteroids are involution forms, and has 
expressed the opinion that they are to be considered as sporangia. Hartleb’s conclusions agree with 
this (1900), and a similar idea is expressed by Winkler who termed the bacteroids, bacterioplasts. 
Hiltner undertook to determine the nature of the injurious action exerted by the substance con- 
tained in the seed-coat of legumes. ‘To this end nodule bacteria were cultivated in water in which 
seeds had been soaked and to which various nutrient materials had been added. 
The clear yellowish liquid obtained by soaking peas for 24 hours was filtered, and clouded 
slightly when an equal volume of alcohol was added, giving a slight precipitate. Tests for tannin 
gave negative results. Corrosive sublimate caused a flocculent precipitate in a 50 per cent alcoholic 
solution. A characteristic precipitate formed when to the solution were added three parts of alcohol 
and a concentrated platinum chloride solution, or, first, platinum chloride and then the alcohol. This 
precipitate was also obtained in the solution after it had been heated for 20 minutes at 120°. A 
second watery extract from the already soaked peas, which was also somewhat yellowish did not 
give this reaction. 
Cultures were then made by putting a loop from a 3 to 4 weeks old gelatin culture into sterile 
water and transferring 0.5 cc. of this to 10 ce. of concentrated watery extract from lupin seeds and 
also into a threefold dilution of the same. ‘These bacteria were taken from nodules on Vicia villosa. 
A test after 24 hours with carbol-fuchsin showed that bacteroid formation had begun in greater 
amount in the dilution than in the concentrated solution. After 2 days in both solutions a rapid 
growth of exclusively rod-shaped bacteroids began. Check cultures in tap water contained practi- 
cally only normal bacteria. In 13 days the concentrated solutions were very heavily clouded and 
had formed a slimy precipitate. In the dilutions there was a somewhat weaker growth. In both 
there were only bacteroids. These were large with numerous vacuoles. 
Bacteria cultivated from pea nodules and locust nodules gave a similar growth in 2 per cent 
solution, but the bacteroid-like bodies which varied greatly in form and in vacuole formation and 
thus suggested involution forms, often seemed to break up into fine granules. No multiplication was 
observable. This modification is traceable to the excretion from the seed-coat. These he regards as 
true involution forms. 
In concentrated solutions, however, and in dilutions to which bouillon was added, true bacteroid 
formation occurred (with pea bacteria), the individuals were all alike, and there was no suggestion 
of a pathological origin. 
Water in which very hard-shelled locust seeds had been soaked did not cause bacteroid formation 
with locust bacteria as had been expected. The water from soja-bean seeds caused a greater forma- 
tion of bacteroids from soja-bean bacteria than did the water from peas upon pea bacteria. 
From his experiments Hiltner concludes that several substances are present in the extract from 
the seed-coats: a pectin-like substance to which the formation of bacteroids is due, and which, instead 
of injuring the nodule bacteria, aids their growth; a further albuminous substance which is likewise 
favorable to the bacteria; third, an injurious substance excreted much sooner than the others in the 
form of a potassium salt which may be precipitated by platinum chloride and which with other 
