1889.] Tubercles on the Roots of Leguminous Plants. 439 



my suggestion is right or not ; at the same time he fully agrees 

 with me and others in believing that these tiny bodies must be the 

 infecting agents, easily and abundantly distributed as they are in the 

 soil, water, &c. 



In the 'Botanisches Centralblatt,' vol. 39, No. 12, 1889, p. 356* is 

 a second paper by Prazmowski, on the nature and biological signifi- 

 cance of the root tubercles of the pea, in which he sums up his views 

 so far. He says, the root tubercles of the Papilionaceaa are not 

 normal structures, but are caused by a special fungus, which inhabits 

 the tubercles, and the spores of which must also occur in the soil. 

 Hitherto he had been unable to determine the true nature of this 

 fungus, but only to show that it penetrates through the root-hairs 

 into the young root, grows in it in the form of more or less branched, 

 unseptate tubes, which are clotbed by a dense refringent membrane 

 and contain innumerable extremely minute rod- like bodies. Under 

 the influence of this fungus the young tubercle is developed in the 

 deeper parts of the cortex, and in its tissues the bacterium-like 

 contents of the fungus become distributed, and grow, divide, and 

 branch at the expense of the protoplasmic contents. He regarded the 

 phenomenon as one of symbiosis, and as benefiting the host as well 

 as the parasite. 



Prazmowski then refers to the papers by Vuillemin, Beyerinck, 

 Hellriegel, and Wilfarth, and says that these instigated him to take 

 up the matter again, and to confine his attention to the pea. The 

 summary of his new results runs as follows : — 



The tubercles are not formed in sterilised media unless infected. 



The infecting organisms are bacteria, identical in form and pro- 

 perties with those cultivated by Beyerinck from the tubercles of 

 various species. From young tubercles the bacteria can be obtained 

 and cultivated pure, and infections from the cultures cause the 

 tubercles to develop. 



The development of the tubercles is only possible in young roots or 

 rootlets ; infection does not occur in older portions of the root system. 



The tubercle-bacteria penetrate through young (not suberised) cell 

 membranes into the root-hairs and epidermis cells of the root, and 

 there multiply at the expense of the protoplasmic cell-contents. 

 Their further development has so far been observed only in root-hairs. 

 After accumulating and multiplying in the root-hair, they unite in 

 racemose colonies at or near its apex ; these colonies become denser 

 and closer, surround themselves with a resistent bright membrane, 

 and join by its means the cell-membrane of the root-hair. Thus 

 arises at the apex of a hair, and on its inner side, a bright knob, 

 usually surrounded by free colonies of bacteria — i.e., colonies not 



* This paper is quoted from the ' Berichte a. d. Sitzungen der K. K. Akad. d 

 Wiss. in Krakau,' June, 1889, and it appears in several journals in the same form. 



