108 BACTERIA IN RELATION TO PLANT DISEASES. 
seem to me doubtful. I think he was experimenting with mixed cultures. Especially do I think his 
theory of alternation of generations, in which Oospora is one stage, and an endospore bearing bacillus 
_another stage, not well supported. Possibly, therefore, the nitrogen stored in his flasks may have 
been due to some other organism than Bact. leguminosarum. Inno part of his paper are the details 
of his experiments so stated that one could reproduce them. Apparently he did not make use of 
poured plates, but depended for isolation on streak-cultures, made in tubes of slant gelatin. 
He states that the nodule-producing organism is pathogenic for some species of animals, e. g., 
rabbits, but this also seems to me not well established by his experiments, since he obtained Oospora 
and an almost round form of very small diameter from the rabbits inoculated with a supposed pure 
culture of the nodule organism. Abscesses formed, locally, in the inoculated animals. 
In 1899, Mazé published a fourth paper on the bacteria of leguminous root-nodules in which 
he reviews the methods of Salfeld and of Nobbe for inoculating the soil with these bacteria, and in 
addition gives some of his own experiences. 
Concerning Nobbe’s work he says: 
To justify the method which he recommends, Nobbe starts out with the following hypotheses: 
There exist in the soil, neutral forms, capable of forming tubercles on the greater part of the 
Leguminosae, and forms adapted to definite species. In general the infection of plants takes place 
by the former, especially in uncultivated soil or in soil which has not borne Leguminosae in a long 
time. The neutral form is modified profoundly by a passage through a leguminous plant becoming, 
in this way, incapable of infecting other species. 
Bacteria thus adapted constitute a definite race: Thus the species Bacillus radicicola (Beyerinck) 
or Rhyzobium pasteurianum (Laurent) comprises a certain number of races each possessing the ability 
to infect particular species of Leguminosae. Sometimes a race is able to attach itself to different 
plants, closely related botanically, but it is not able to utilize atmospheric nitrogen upon these 
inappropriate hosts. 
Mazé then raises the question, not mentioned by Nobbe, as to how these races pass over from 
one season to another. He says: 
“May we conclude that they retain after months and years the ability of their ancestors to 
live incapable of attaching themselves to plants of other species than the one which previously 
sheltered them? Nothing would be less justifiable than such an assumption. It has long been 
known in bacteriology, that all species of bacteria are subject to the influence of the medium on 
which they live. More than any others, the bacteria of the Leguminosae possess this adaptability 
which assures the dissemination and preservation of a species.”’ 
Mazé claims that forms living in the soil lose, little by little, the characteristics which made 
them easily identified when taken directly from the nodules. A dilution of soil applied to plants 
growing in nutrient solutions caused nodules after 15 days. The same dilution inoculated on a series 
of agar tubes, made for the purpose of obtaining isolated colonies, did not give any forms which 
corresponded either morphologically or physiologically to the typical bacterium of the nodules. By 
a long series of passages with all the species obtained from these cultures he states that finally two 
forms were obtained which he identified by inoculations as the root-nodule organism. From this he 
draws the conclusion that forms isolated from the soil acquire gradually, when subjected to a medium 
containing the proper carbohydrate and nitrogen, the ability to elaborate the mucilaginous substance 
and to fix atmospheric nitrogen. He thinks, therefore, that this ability is very unstable with the 
bacteria of the Leguminosae. They acquire it in the nodules and lose it in the soil. 
He gives the following experiment as proof of this: 
He sowed the nodule bacteria on both sterilized and unsterilized soil kept saturated during 
the whole experiment. On the unsterilized soil, conditions favored the growth of soil bacteria. 
At the end of 8 months it was impossible to obtain colonies resembling those which supplied the 
bacteria sowed. On sterilized soil the bacteria removed from competition with other soil bacteria, 
retained their initial characteristics after 8 months. 
He states also that the characteristics of these races of bacteria at the moment of isolation from 
the nodules are far from being as distinct as Nobbe claims. Thus, for example, a bacterium coming 
from one leguminous species is capable’ of attaching itself to certain other species. Nobbe admits 
this but thinks that while able to form nodules the bacterium is no longer able on these strange 
plants to fix nitrogen and so it becomes a parasite which is frequently injurious. 
Mazé does not agree with this last statement: Bacteria from any of these plants will fix nitrogen 
if they have sugar and enough initial nitrogen. The plants all offer that, and he says that the only 
condition requisite to nitrégen fixation is their ability to penetrate them. This ability, he thinks, 
depends on the alkalinity or acidity of the soil. He found that lupins inoculated with bacteria from 
furze and broom formed just as many and as large nodules as those inoculated with bacteria from 
the lupin, while the checks showed no trace of nodules. The furze and broom came from uncultivated 
