1887.] Roots of the Leguminosece. Abrus precatorius. 331 



VI. " The Tubercular Swellings on the Roots of the Legu- 

 minoseas." By H. Marshall Ward, M.A., F.L.S., Fellow 

 of Christ's College, Cambridge, and Professor of Botany 

 in the Forestry School, Royal Indian College, Cooper's Hill. 

 Communicated by Prof. M. Foster, Sec. R.S. Received 

 April 25, 1887. 



(Preliminary Note.) 



The author finds that the tubercles on the roots of the Legu- 

 minoseae are due to the action of a parasitic fungus. Not only has he 

 produced the tubercles by infection from without, but he has also 

 found the infecting agent, and repeatedly seen and figured the infect- 

 ing hypha passing down inside a root-hair and across the cortex of bhe 

 root into the young tubercle. Here the hyphal branches bud off 

 yeast-like cells, which are extremely minute and numerous, and 

 resemble bacteria at first sight ; they differ in their mode of multi- 

 plying by budding. 



The action of these minute germ-like bodies causes the protoplasm of 

 the cells of the root to assume plasmodium-like characters, and 

 induces the flow of nutritive substances to these cells, and hypertrophy 

 results. On the decay of the tubercles, the germ-like bodies pass 

 into the soil (where they can always be found) and infect other roots ; 

 it is very probable they may be of extreme importance in agriculture. 



VII. "TheProteids of the Seeds of Abrus precatorius (Jequirity)." 

 By Sidney Martin, M.D. Lond., Fellow of University 

 College, London, and Pathologist to the Victoria Park 

 Hospital. Communicated by Prof. E. A. Schafer, F.R.S. 

 (From the Physiological Laboratory, University College, 

 London.) Received April 21, 1887. 



The proteids of the seeds of Abrus, the Indian liquorice, are im- 

 portant physiologically, because they have been shown (by Warden 

 and Waddell*) to be possessed of poisonous properties. To the 

 poisonous product extracted by these observers the name " abrin " 

 was given ; and though it was decided that abrin was closely allied to 

 " plant-albumin," yet no experiments were recorded to show whether 

 the product was a mixture or a single proteid. They obtained it by 



* 4 The Non-bacillar Nature of Abrus-poison.' By C. J. H. Warden and L. A. 

 Waddell. Calcutta, 1884. 



VOL. XLII. 2 B 



