R42 
gummy exudation, This watery solution can contain but two potent 
clements—the bacteria and the accompanying soluble substances— 
to one of which the results of the inoculation must be due. If the 
disease is brought about by the deleterious substances which are the 
product of the organisms activity, as some have claimed for zymotic 
diseases of animals, then such an inoculating solution, when freed 
from all bacteria, ought still to be capable of starting blight, 
although the disease might not progress very far without the pres- 
ence of specific bacteria to supply the necessary quantity of the poi- 
son. To test this point, on July 18, a strong solution made from 
blighted tissue was filtered through a new porous cell, such as used 
for small electric batteries (in this and other experiments needing a 
considerable quantity of material freshly blighted, green fruit was 
used which had been artificially inoculated). The filtrate was color- 
less and free from all bacteria; it was used (Exper. 378) to inocu- 
late green fruit (which show results more surely and quickly than 
branches) of Bartlett pear, and at the same time another fruit on 
same tree was inoculated with the unfiltered liquid (Exper. 379). 
In a few days the latter was filled with the peculiar viscid products 
of active blight, showing that the unfiltered liquid had the elements 
of disease in it; the former, however, remained wholly unaffected, 
and a month afterward was still perfectly healthy, having only a 
slight scar where the inoculation had been made. 
Not to rest so important a matter upon a single trial, however 
conclusive it might seem, essentially the same experiment was repeated 
on July 24. Another porous cell was used, and the filtermg done 
as before. This time two Bartlett pears were inoculatea with the 
unfiltered solution (Exper. 387 and 388), and two with the filtrate 
(Exper. 389.and 390). Both the former gave signs of blight in 
three days, by the tissues turning brown for a quarter inch around 
the puncture; in two days more this had extended to nearly an inch, 
and No. 387 was removed for microscopic examination. No. 388 
was left on the tree six days longer, turning brown nearly through- 
out; upon cutting open, only the stony and other solid tissues 
remained, the soft tissues having been changed into liquid by action 
of the bacteria, and this nearly all escaping at the opening made 
by the puncture left an unoccupied cavity. But, as in the former 
trial, the inoculation with the filtrate produced no effect. This is 
accepted as a complete demonstration that the juices accompanying 
blight will not induce the disease in any form when freed from the 
blight bacteria. 
Having shown that the juices accompanying the bacteria are inca- 
pable of producing the blight, it is desirable to show what action the 
bacteria will have when thoroughly isolated from all substances, 
solid and liquid, with which they are usually associated during the 
active stages of the disease. This was accomplished by a series of 
artificial cultures in an infusion of corn meal. The infusion was 
prepared by boiling corn meal with water, and filtering. The color- 
less nutrient liquid thus obtained was placed in various forms of 
culture vessels, and sterilized by heating. There was introduced 
into one of these vessels, in March, a tiny fragment of wood from 
the interior of a limb of Flemish Beauty pear, in which the disease 
a 

