conspectus: action on the plant 45 



repetition of this experiment gave the same results — good growth 

 of checks and astonishingly slow growth of plants from the 

 heated seed. 



We are now (1918) growing Ardisia plants from surface- 

 sterile seeds in flasks of glowed sand in nitrogen-free media. ' 



The bacterial nodules on the leaves of the East Indian 

 Pavettas (first described by Zimmermann) are apparently of a 

 similar nature, but here the bacterial foci are scattered over 

 the surface of the leaf, often, however, with astonishing regu- 

 larity (Figs. 30 to 34). 



Fig. 34. — Bacteria from a nudulr >>u leal of Pavetta angustifoUa. X 1000 

 Stained witli anilin blue. 



The action of such organisms as I have just mentioned differs 

 probably from the behavior of active parasites in that they 

 liberate much weaker toxins and enzymes, can attack only very 

 actively growing parts, and also give off compensating nitro- 

 genous substances. Not yet proved for Ardisia (1915) but 

 proved apparently for the Pavettas by Dr. F. C. von Faber whose 

 paper is in Jahrb. f. wiss. Bot., 1914, Bd. 54, p. 243. ^ 



1 Dec, 1919. These plants have remained alive for 18 months, but the foliage 

 is paler and the plants have made less growth than checks in ordinary soil. The 

 experiment is not conclusive because at the end Dr. Jodidi found a trace of 

 nitrogen in the sand (0.5 mg. per kilo). 



2 Since this was written Miehe claims to have proved it for Ardisia, but owing 

 to the war I have not been able to obtain his paper, which is in Ber. d.d. Bot. Ges., 

 xxxiv Bd., 1916, pp. 576-580. 



