10 BACTERIA IN RELATION TO PLANT DISEASES. 
Under (a) there should be numerous observations on many plants, with very 
careful microscopic examination of stained and unstained material. The cells of 
many plants contain granules which often dance about so actively (pedesis or 
Brownian movement) as to be very deceptive, and yet they are not bacteria. Living 
bacteria in plant tissues can always be stained so as to stand out distinctly if the 
sections are well prepared and sufficiently thin. When bacteria occur in plants as 
parasites they are usually very abundant in the vascular system, or the parenchyma, 
or both, and there is, so far as yet known, always a distinct breaking down (solution) 
of some portion of the tissues (see figs. 6 and 7, and plate 3). If the parenchymatic 
tissues are sound, if there is no bacterial ooze on making sections, if the vascular 
system is not occupied, and if bacteria can not be demonstrated in the tissues by 
proper staining, then it is very unsafe to infer their existence from dancing particles, 
no matter how many may be visible in the unstained sections. Moreover, bacteria 
may be present in some of the plants and not in others, z. e., not constantly present, 
and so not the cause of the disease. It is conceivable that they might also be present 
*Fic. s.—Bacterium campestre parasitic in a turnip-root (inoculated plant No. 53). This figure 
shows the bacteria crowding out into the cells surrounding the reticulated vessels. The tignified 
portion of each vessel is indicated by fine dots. Material fixed in strong alcohol, infiltrated with 
paraffin, cut on the microtome, stained with carbol-fuchsin, and tthe excess of stain removed in 
dilute alcohol, section then dehydrated and mounted in xylol-balsam. Drawn from a photomicro- 
graph, the contrast here indicated being not greater than that shown in the section. X 500 circa. 
