ROOT-NODULES OF LEGUMINOSAE. 135 



The morphology of the bacteria varied with the species of legume, the age and size of the nodule, 

 the portion of the nodule examined, and the conditions of infection and growth. In plants of the 

 tribe Phaseoleae the bacteria were mostly small rods with comparatively few branching and irregular 

 cells. In plants of the tribe Trifolieae branching and irregular forms prevailed. In general, although 

 not always, simple rods prevailed in young and small nodules and branched irregular forms in older 

 and larger nodules. The proximal part of the nodule, that part first formed, may contain simple 

 rods mainly, and the distal part, where growth is taking place rapidly, may contain not only simple 

 rods but many branched or twice branched forms. 



The authors state that they were able to see the polar flagellum of this organism unstained by 

 taking a loop of the mucilaginous or viscid growth of an agar culture a few days to several months 

 old and spreading it out in streaks on a clean slide, lashing it out into slender tongues. The film was 

 allowed to dry in the air but not killed or fixed in any way. It was then flooded for a moment with 

 a saturated alcoholic solution of gentian violet to stain the mucilage, washed under the tap, dried 

 between folds of filter paper, and examined with an oil immersion lens. The mucilage stained deeply, 

 and the flageUa not at all. On the margins of this slime they frequently found the flagellum visible 

 as a long colorless thread attached to the pole of the bacterium.* Kiskalt's amyl-Gram stain is stated 

 to be very good for staining Bacterium leguminosarum, especially from the nodules, since the amyl 

 alcohol clears up the background, without removing the stain from the bacteria, and also shows 

 clearly the internal structure of the bacteria. The authors give a table, which includes the length 

 of life of Bacterium leguminosarum from various legumes (white clover, red clover, alfalfa, vetch, flat 

 pea, common pea, bean, Desmodium, and Apios). They found this to vary from i to 2 years when 

 cultivated on soUd media containing wood ash, maltose, acid phosphate, and agar. In most instances 

 the organism lived more than 18 months, once it lived 2 years lacking 5 days; in 2 cases only was 

 it dead at the end of 10 months. 



The authors tried various methods of sterilizing the surface of seeds, viz., dry heat, moist heat, 

 sulphuric acid, calcium hydrate, formalin, and mercuric chloride, but only in a few instances were 

 living seeds obtained free from living bacteria, and never when the seeds were inoculated with the 

 spores of bacteria. Either the seeds refused to germinate after treatment or else after germinating 

 they were found to be still infected with bacteria. The authors finally obtained sterile seeds by 

 another method, viz., by selecting well-developed, unopened pods, soaking these for an hour or more 

 in I :i,ooo mercuric chloride water, and then placing them in folds of sterile cotton. The pods were 

 also held in forceps, and passed through the flame on all sides and the ends were well burned. They 

 were then opened, the seeds removed and placed between folds of sterile cotton. In a few days, 

 when the seeds were dry, they were taken in flamed forceps and put into plugged sterile test tubes 

 where they remained at room temperature until they were needed. In this connection see fig. 2. 



To germinate the seeds, they were thrown into about 3 cc. of boiling distilled water in test tubes. 

 The tubes were immediately cooled and incubated at 37° C. and then at 25° C, until the seeds 

 germinated. After the first 24 hours the tubes were inclined so that the seeds were only partially 

 under water. The germinated seeds from those tubes in which the water remained free from bacterial 

 growth were then transferred to flasks containing the ash-maltose-agar, where the growth of the 

 seedlings continued. 



At first all of the plants grew vigorously on the agar, but in those flasks held as checks they 

 began to dwindle after a time and either died outright or made a very unsatisfactory growth; whereas 

 in those flasks which were inoculated with Bacterium leguminosarum the growth continued until the 

 end of the experiments. One series of flasks was under observation for 8 months and three other 

 series for shorter periods. The root-nodules began to appear in about i month. Sometimes they 

 were few and large; at other times many and small. In one instance 70 developed on the roots of 

 one plant. In the absence of root-hairs the infection began as a small transparent spot in the root. 

 Nodule formation and general infection of the root appeared to check extensive root formation. Bac- 

 terium leguminosarum grew copiously on the surface of this agar and penetrated the agar wherever 

 the roots penetrated it. A general infection of the roots usually accompanied the nodule formation. 

 In flasks with ash-maltose-agar, nodules were obtained upon Phaseolus vulgaris, Vicia villosa, Pisum 

 sativum-, and Glycine hispida. Cross-inoculations were not observed; for instance, when pea and 

 vetch were grown in the same flask and inoculated from a vetch culture, nodules formed only on the 

 roots of the vetch. In the same way, when a bean and pea were grown in the same flask, inoculated 

 from the pea culture, nodules formed only on the roots of the pea. 



One can not tell from this paper whether or not the organisms used for inoculations were sub- 

 cultured from colonies sufiiciently to overcome de Rossi's objection. Barlow with whom I conversed 

 thinks some of them were. 



*The writer has examined a slide of this kind prepared by Mr. Barlow. Only here and there does a rod show this 

 phenomenon, most do not. 



