258 



SCIENTIFIC AGRICULTURE 



June, 1921. 



to the rather short description given by Beijerinek. 

 only the more complete one published by the senior 

 author in 1905 exists thus far. On account of its 

 great similarity to B. radicicola. B. radiohacter should 

 be very well" known to all bacteriologists working 

 with the nodule bacteria in order to avoid mistakes 

 whic.h may otlienvise not be discovered until only 

 negative results are obtained in the inoculation tes-ts. 



Concerning the flagellation of the nodule bacteria 

 three statements have been published more recently 

 which also will have to be discussed presently. Ac- 

 cording to J. K. Wilson (18) the soybean bacteria, are 

 peritrichous : BartHicl (2) declared lupine and alfalfa 

 bacteria to be lop.hotrichoiLS; Fred and Davenport (7) 

 found the alfalfa organism peritrichous, but they 

 found the lupine bacteria characterized by having one, 

 rarelv two, flagella. 



EXPERIMENTAL RESULTS. 



The followuig strain.s of nodule bacteria were stud- 

 ied after having been tested witlh positive results in 

 regard to their ability to produce nodules on the 

 ilio.st plants from which t.tiey were isolat^^d. 



1. Cowpea. fi. Red clover. 



2. Peanut:. 7. Sweet clover. 



3. Japan Clover. S- Vetch. 



4. Beggar "Weed. 9- Strophostyles. 



5. Soybean. 

 There were also included in our investigations two 



.^trains isolated from: 



10. Black locust, 

 n. Lupine. 

 No positive inoculation test could be made on blad- 

 locust. The lupine culture was kindly furnished bv 

 Dr. E. B. Fred, of the University of "Wisconsin, who 

 had tried it with positive resiilts on this plant. Our 

 tests were equally successful. 



Two noninfectious "fast erowing" cultures isolate l 

 from lesume nodules and identified as Barilhix rarJm- 

 hacter were studied in comparison with six Radio- 

 • bacter strains which originated from soil and which 

 were kept in the senior aut'nor's collection since the 

 years given in parent'heses. 



12. Fast grower from 15. Same (1907). 



cowpea. 16. Bacillus radiohacter 



13. Fast grower from from soil (1908). 



soybean. 17. Same (1908). 



14. Bacillus radiohacter 18. Same (1908). 



from soil (1904). 19. Same (1916). 



No. 14 is the strain which in 1904 had been acknow- 

 ledged by Prof. Beijerinck as being identical with his 

 Bacillusradiohactrr and which was used by the senior 

 author for the original dcsicription published in 1905 

 (12). 



The results of our investigations leave no doubt t.liat 

 tihe earlier findings of the .iunior author were correct 

 so far as tlie polar flagellation and the peculiar 

 morphological and cultural features of the cowpea- 

 soybeau organisms are concerned. On the other hand, 

 it could now be ascertained with equal certainty that 

 the bacteria producing nodules on clover, alfalfa, 

 veteh, and ot.aer plants originally cultivated in 

 Europe are all peritrichic and exliibit all tfhe charac- 

 teristics of BaciUux radicicola. as described by Bei- 

 jerinck and other authors. 



Those findings which were obtained most frequently 

 and which may be considered as being typical for the two 

 groups of nodule bacteria are compiled in Table I, 



liogether with analogous data pertaining to Bacillus 

 radiohacter. Photographs of the most characteristic 

 detaibi are reproduced on Plates 68 and 69. 



Wlien grown from the root nodule on Harrison and 

 Barlow's ash agar, mannite agar, or similar agar of 

 low nitrogen content, the two groups of nodule bac- 

 teria are best characterized and differentiated by the 

 ver>' slow growth of colonies in t.he cowpea-soybean 

 groii]] and the comparatively c|uick growth of tho.se 

 of Bacillus radicicola. Frequently, Init not always, the 

 devel(i]iineiit of B. radiohacter is .still somewhat more 

 rapid than that of B. radicicola ; in the macroscopieal as 

 well as in the microscopical aspects, liowever, the colonies 

 of these two species on such media are so very much alike 

 that it is almost impossible to distinguish them with cer- 

 taintv. Both, when develoning on the surface, are per- 

 fectly round, drop-like. soft, waterv or slimv. glistening, 

 tran.sparent. Often a whiti.sh center or whitish .streaV*-- 

 become visible within the more transnarent mass, es- 

 np(^ia11v if the .surface colony is t|he outgrowth "f «" 

 iinKorlrled colony. Sometimes it appears as if t^^" 

 wliitish center werp resmlarly to be seen onlv ■"•Ub 

 certain strains of Radicicola and Radiohacter. T'i'S 

 is "Oil the ease, however. Its presence or absencp is 

 erratic and cannot be used for differentiation. The 

 imbedded colonies are always small, white, opariue. 

 mostly lentiform. less frecniently circular. TTnder the 

 microscope the surface colonies present tihem.selves as 

 Sharp-edsed disk-^, pure white at the outside with 

 yellow-grayish granulation in fhe center. In a few 

 cases a radiate .structure becomes visible. The colonies 

 of the cowpea-soybean srrouD appear macrascopicallv. 

 as well as microscopically like young colonies of the 

 Radicicola type. The presence of Radiohacter colonies 

 on the plate stimulates their growth markedly. 



In cell morp.nolog\' there is again a more pronounced 

 relationship between Radiobacter and Radicicola than 

 between the nodule bacteria of the clover-vet<')Ii group 

 on the one side and of the cowpea-soybean group on 

 the other. This holds true with the regular rod fonns 

 as well as with t.he very pleomorphic, curved, swollen, 

 branched, or otiherwise cfhaiiffed types of growth char- 

 acteristic of these groups. The photographs on Plate 

 68, D-L, represent tjie pictures we have seen most 

 frequently, but they do not pretend to give a com- 

 plete illustration of the wide pleomorphlsm of fliese 

 organisms. Before their complete life hi.story can be 

 given much additional material will have to be col- 

 lectied, especially with regard to the form of goni- 

 dangia, regenerative bodies, and the various cells 

 developing from the .symplastic stage. At present 

 we intend only to bring outi as clearly as possible 

 those points whic.ii are important for a correct dif- 

 ferentiation and diagnosis. As far asi one may judge 

 from the micro.scopic appearance, it is tihe inclination 

 of Bacillus radiohacter to form stars w'hich is most 

 characteristic (PI. 68, L), and this was used, there- 

 fore, by Beijerinck for its denominatiion. "With B. 

 radicicola the frecpient occurrence of the clear-cut, 

 compact Y forms is the mo.st conspicuous feature 

 (PI. 68, H) ; wjtiereas tihe bactieria of the cowpea-soy- 

 l>ean group present themselves in most cases, when 

 stained with aqueous aniline dyes in the usual manner, 

 as i^hort or long, unstained sheaths with one or more 

 darkly stained granules (PI. 68. J). Of coui-se Y 

 forms, as well as unstained sheaths with darkly 

 stained gonidia, can be ob.served not infrequently 

 with the other organisms, too, and t/ie star forma- 



