41 



thick slime was acid to litmus, especiaH^^ when diluted with distilled 

 water. The tissues were softened and there was a peculiar smell, which 

 was not rank or strong. 



The l)ehavior of Ps. stevjarti on this substratum differed from that 

 of Ps. hyacintJd and Ps. campestHs in the same way as on the yellow 

 globe turnip and was even more pronounced, so that it might be used 

 as a means of distinguishing these organisms. The growth on the 

 seventh da}^ was about one-tenth as much as Ps. kyacinthi and one- 

 twentieth or one-thirtieth as much as Ps. carnpestris. 



On the eighteenth day the differences were as follows: 



Ps. stevxvrt'i : Growth, buff yellow, thin, covering the whole of the 

 air-exposed surface, but not dense enough to hide the slight irregu- 

 larities of the substratum (not smooth). Surface slightly iridescent 

 and with fine striaj (Zeiss X 6 aplanat), precipitate buff yellow and 

 moderate in amount, water not grown full of the solid 3'ellow slime, 

 substratum not browned. 



Ps. hyacinthi and Ps. carnpestris: Slime in the air copious, smooth, 

 ver}' wet-shining, pale yellow, surface not iridescent. Fluid grown 

 full of the yellow slime (solidified). Substratum browned or read}^ to 

 brown. 



In old cultures of Ps. stcirarti there was no increased growth, no 

 brown stain, and no softening of the tissues. On the thirty-fourth 

 da}' the thick slime would not wet litmus paper until water was added, 

 when it gave an alkaline reaction. On the sixty-fourth day there was 

 "a peculiar smell" and a feebly alkaline reaction. The iridescence 

 persisted. 



Carrot. 



Cylinders of carrot were prepared in the same wa}' as the potato 

 cylinders. 



Ps. hyacinthi grew well on this medium at 20° to 23° C. , and gen- 

 erally without any distinct retardation. Usually growth was visible 

 on the third day, and continued for several weeks, covering the aerial 

 part of the cylinder with a bacterial layer a millimeter thick. The 

 fluid in the bottom of the tube (1 to 2 c. c.) was also filled with a thick 

 yelloAV slime, so that after 3 weeks it could usually be turned bot- 

 tom up without flowing. Generally, though not always, growth was 

 copious enough by the end of the first week to obscure the oi*ange red 

 of the substratum, which was not the case with Ps. camjjcstris. 



The surface was always wet-shining, even in very old cultures. In 

 some it was smooth and homogeneous-looking from the start, and 

 remained so. In others the surface was shagreened at first, but after 

 eight days became smooth and homogeneous-looking. The bacterial 

 slime was not sticky on the eighth da}', in which particular it is very 

 unlike Bacilhis tracheiphihis. Subsequently (thirty-first and sixty- 

 seventh days) it became slightly sticky. 



