1889.] Tubercles on the Roots of Leguminous Plants. 433 



identity of the bacteroids in the two cases. In some of the cultures 

 made in the summer of 1888 I infected the roots of the pea with bacte- 

 roids taken from the tubercles of the bean, and as this is a point 

 of some importance, in view of the belief that each species of Legu- 

 minosse may have its own species of bacteroid, I may say a few words 

 on this phenomenon. 



Having satisfied myself that when the contents of the tubercles of 

 the bean are placed among the root-hairs of the young bean root, the 

 latter become infected, I prepared a number of beans and peas as 

 follows : They were allowed to germinate in the sterilised sand until 

 the radicle was about half an inch long ; each seedling was then 

 pinned to a cork, and so fixed that the radicle pointed downwards 

 into a large wide-mouthed bottle, in which the cork fitted. The 

 bottle was then carefully lined with filter-paper, kept moist by dis- 

 tilled water ; the cork fitted closely, and thus the atmosphere in the 

 bottle was sufficiently damp to enable the radicle to go on growing 

 at the expense of the reserve materials in the cotyledons, and in 

 course of time to put forth a dense pile of delicate root-hairs. As 

 each seedling was pinned on to the cork, I sprayed on to the surface 

 of the radicle, by means of a freshly-drawn capillary tube, a mixture 

 of bacteroids and water made as follows : Tubercles of the bean-roots 

 were carefully washed, placed in alcohol for a few minutes, and then 

 fired, then again washed with distilled water, and pounded in a 

 mortar with distilled water : small drops of this were placed on the 

 radicles as said — i.e., on the radicles of both peas and beans. 



In other cases I employed the hanging drops in which I was 

 attempting to cultivate the bacteroids. These consisted of nutritive 

 solutions with asparagin, and with or without gelatine, and in which 

 were placed a few of the bacteroids obtained from cleaned tubercles 

 (cut with a razor sterilised by heat) by means of sterilised needles. 

 I may here say that these" cultures {i.e., as micro- cultures) have given 

 me much trouble, and little results: to obtain pure cultures is a 

 matter of greater difficulty than Beyerinck's paper would lead one to 

 expect, and it is not proposed at present to Jay much stress on 

 the evidence got from them. Nevertheless, colonies are obtained, and 

 in some cases at least I have transferred the infecting organism from 

 these cultures to the root-hairs of peas and beans. 



In any case, I have succeeded in obtaining extracts of the tubercles 

 which contain the infecting germs, and although the latter were 

 always taken from the tubercles of the bean, they infected the root- 

 hairs of both peas and beans equally well. 



It is especially the very young root-hairs, with extremely delicate 

 cell- walls, that are infected, and the first sign is the appearance of a 

 very brilliant colourless spot in the substance of the cell-wall (figs. 

 A and B) : sometimes it is common to two cell-walls of root-hairs in 



2 g 2 



