1881.] Bacteria as a Cause of Disease in Plants. 529 
tion of the starch bacteria progresses. How these originally 
gain entrance to the cells was not made out. There are cer- 
tainly no pores or other visible openings through which they 
pass. Water, however, is absorbed by the cell walls, and passes 
through their molecular spaces in all directions. It may be that 
in the germ condition the bacteria are really small enough to 
pass with the water through the walls, notwithstanding the fact 
the highest powers of our microscopes fail to detect the molecu- 
lar openings. However this may be, it is positive enough that 
the adult bacteria do not in this way traverse the cell walls. 
The evidence is totally against any distribution of the organisms 
in the tissues by the circulations of water or sap. They slowly 
make their own way from cell to cell, progressing equally in all 
directions from the starting point when the same conditions are 
presented. 
On July 1, 1880, I inoculated two pear trees by inserting small 
pieces of bark from a pear tree in which the disease was in active 
progress. On the 12th and thereafter, inoculations were made by 
dipping a clean needle or the sharp point of the blade of a pen- 
knife into the viscid substance exuding from diseased bark, dilu- 
ted or not with distilled water, and thrusting the instrument into 
the experimental trees. Usually three such punctures were 
made near each other, but the three were counted as one inocula- 
tion. 
No visible results followed any of these inoculations during 
the first eight days, and in some cases for two or even three times. 
this period. In the majority of cases ten to twelve days elapsed 
before external signs of the disease could be observed. No differ- 
ence was detected in this or any other respect in the different 
Ways of inoculation. But numerous external applications of the 
virus were made to the bark and leaves without wounding, none of 
Which seemed to communicate the disease. The trees were ex- 
amined at least once each day until the 14th of August, 1880, 
and every observed change carefully noted. Very often the dis- 
* cate could be detected by dissections and microscopical examina- 
tions when no external indications were presented, but these were 
Rot counted as successful inoculations except in very clear cases. 
The €xperiments upon pear trees were made upon trees three 
years old, Bartlett and Clapp's Favorite, seventy-two in num- 
ber. The apple trees are Grime’s Golden, and the quince, Angers. 
VOL, XV.—nNo, vir, 37 
