39 



this color resembled tawny olive, but was paler. On the thirty-second 

 day it was still only a pale ])rown. After 69 days the color of this 

 pioment was between Rido-way's russet and ])urnt umber. 



Bacillus cmiylovor lis made only a moderate growth on this substratum, 

 and produced no brown stain, but developed an acid. 



After a year or two this test of Ps. JiyiainfJii was repeated at '10 to 

 25^ C. , using- yellow globe turni|)s (a rough-leaved, nonglaucous sort). 

 On the third day 5 s([. cm. of the slant surface was covered with a 

 smooth, wet-shining slime, which was abundant enough to hide the 

 substratum. On the seventh day there was a copious 3'ellow, smooth, 

 wet-shining growth over the whole cylinder and in the water, but no 

 browning of the substratum. On the eighteenth da}' the fluid was so 

 full of the yellow slime that it would notiiow when turned bottom up, 

 and there was a slight browning of the upper part of the substratum. 

 On the twenty-seventh day there was a distinct pale-brown stain in the 

 upper part of the substratum. On the thirty-fourth day the slime 

 was neutral to neutral litmus paper. On the fiftieth da}' the color was 

 between ])urnt umber and mununy brown, and the fluid was grown solid 

 with the yellow-brown bacteria. On the sixty-fourth day the color of 

 the substratum was burnt umber. The culture had a faint, peculiar 

 smell. The outline of the substratum was preserved, but on being 

 removed from the tube it was nmshy soft to the fingers, and even to a 

 piece of litnuis paper which could be thrust into it. The substance 

 was feebly alkaline throughout. There were some involution forms, 

 but nothing resembling spores. Large crystals were present. 



No starch remained, if any was originally present. The middle 

 lamella was dissolved or greatly softened. The cell wall proper (of 

 the turnip) was apparently intact, but for the most part the contents 

 of the cells were gone, although some large and small rings of doubtful 

 origin remained. With Russow's cellulose test many of these cells of 

 the substratum did not stain at all, a few became deep blue, and a few 

 deep purple. In most, the walls remained colorless, but the contents 

 of the cells reacted pale blue. Corresponding results were obtained 

 with chlor-iodide of zinc. The contents of the cells frequently became 

 ))lue while the walls remained colorless or turned to brown or reddish 

 brown. Doubt was thrown on these results, however, by the behavior 

 of the check tubes, which also gave an uncertain cellulose reaction 

 with these reagents; i. e., (ell walls purplish in the chlor-iodide of 

 zinc (on long soaking), and bright blue only in a few cells and parts of 

 cells with Russow's test. 



Ps. caiiipestris also made a prompt and copious growth on this sub- 

 stratum, but there were some diHerences. On the seventh day the 

 growth, whil(^ veiy al)undant, was scarcely distinguishable in color 

 from the substratum: i. e., it was plainly less yt^llow than that of Ps. 

 hyacintld. At this date the fluid was grown full of the bacteria 



