130 



other green-fluorescent bacteria) first converts the nitrate to nitrite, 

 and then liberates the azote as free nitrogen.' Gas bubbles were given 

 ofl' continuall}' during the first few days, so that the top of the fluid 

 was foamy, as if it had been shaken violently. During this stage the 

 liquid gave a deep blue-black reaction with boiled starch water, potas- 

 sium iodide, and sulphuric acid. Later the gas bubbles disappeared, 

 and then (sixteenth day) no nitrite reaction could be obtained. The exper- 

 iment with B. pyo. pericardii Idis was repeated, using fermentation 

 tubes; a considerable quantity of gas collected in the closed end. This 

 gas was not absorbed on shaking with caustic soda (absence of CO.^); 

 it did not diffuse quickly or explode when it was tilted into the open 

 end of the tube and a lighted match applied (absence of I13 drogen and 

 marsh gas): lighted matches thrust into the ])owl were repeatedly 

 extinguished (presence of nitrogen). 



One or two other interesting facts were observed in connection with 

 cultures in the nitrate bouillon. /*y. stewarti and B. mnylovorus made 

 a very feeble growth in comparison with Pi<. hyacintJd and P><. cmiLpes- 

 tris. B. coli grew l)etter, throwing down in 16 days about 10 times 

 as much precipitate as B. <iinylov<yru8. In early stages of growth, i. e., 

 during the first 2 or 8 days, the -1 cultures of I\. campestrls were ver}^ 

 different from those of /*s-. hyachitJu in that the former contained 

 many hundreds of tiny white zoogkjea^ scattered uniformly through the 

 liquid, giving it. especially under the Zeiss X 6 aplanat, a distinctly 

 granular appearance. On the sixteenth day this phenomena had dis- 

 appeared and the cultures of the two organisms were then much alike. 



Ps. p>haseoli was also tested in this nitrate bouillon. Like Ps. cam- 

 pestris, it formed great numbers of small zoogloese during the first few 

 days of growth. It was entirel\' unable to reduce the nitrate to nitrite 

 in this solution (14 days). 



FERMENTS. 



No attempt has been made to isolate any ferment, but the behavior 

 of Ps. hyacinth I in the host plant and in various culture media leads 

 to the conclusion that several enzymes are secreted. 



Cytase. 



The thin, non-lignified walls of the spii-al vessels of the host plant 

 are dissolved, letting the bacteria out of the vascular system into con- 

 tact with the parenchyma. Fragments of the spiral threads are 

 apparently all that remain of these vessels in bundles which have been 

 long occupied. Once in contact with the parenchyma, cavities, filled 

 by the l)acteria, are formed in this tissue, the cells being first sepa- 

 rated from each other and finally destroyed, as Dr. Wakker has 

 described. These facts indicate the secretion of a cytohydrolytic 

 enzyme. At the same time the slowness with which the vessels are 



* These are the organisms that reduce the value of the farmer's manure pile. 



