128 
Journal of Agricultural Research 
Vol. XXV, No. 3 
Together with the middle lamellae they were dissolved after treatment 
with dilute acids followed by dilute alkalies. It appears that they are 
composed of some pectic substance, possibly calcium pectate. 
LOCATION OF THE BACTERIA IN THE INTERCELLULAR SPACES 
Evidence has been accumulated which shows that the bacterial granules 
which were observed in the intercellular spaces bordered by yellowed 
walls are the crowngall bacteria in situ. These bacterial bodies have 
been observed to be constantly associated with crowngall tissue in 
tomato stems through four series of inoculations in which the galls were 
examined at two-day intervals from the day of inoculation until the galls 
were 22 days old. At the latter age they showed the characters of 
maturity—that is, they had well-developed hypertrophic and hyperplastic 
areas which were abundantly supplied with new vascular elements. The 
bacterial bodies were found very easily in the early stages and without 
much difficulty in the later ones because of the yellowing of adjacent 
walls which accompanies their presence. This yellowing is similar to 
that which may be observed in connection with various types of injured 
tissue. Consequently, although crowngall bacteria in the intercellular 
spaces appear to be accompanied by a yellowing of the surrounding 
walls, this color does not always indicate the presence of bacteria. 
With the aid of polarized light it was observed that the yellowed walls 
had lost their property of double refraction. It appeared that the 
bacteria had produced some change in the cellulose of adjacent walls. 
However, a further consideration of the nature of this action is beyond the 
scope of this paper. 
The phenomena just described were not observed in either sound 
tissue or tissue that had been punctured but not inoculated. The most 
likely sources of confusion were found to be the granules of pectic sub¬ 
stance already described and deposits of very tiny crystals of calcium 
oxalate. These latter were observed quite commonly in the gall tissue. 
Their identity as crystals was very easily established by the use of 
polarized light. 
After the location of the bacteria was ascertained the method of 
demonstrating them in the paraffin sections became simplified. It 
appeared that the customary methods had failed because the bacteria 
produced an effect on the cell wall that made it take up the stain more 
heavily than did the normal walls. This resulted in a masking of the 
bacteria when they were not present in very large numbers. It was 
found also that the intercellular spaces occupied by the bacteria gave a 
positive protein reaction to Millon’s reagent, while the bacteria themselves 
did not. Whether this substance is produced by the bacteria or by the 
neighboring cells is not understood, but its presence certainly renders 
more difficult the demonstration of the bacteria. 
It remained to discover a combination of stains that would color the 
bacteria and still not hide them by staining too deeply the walls and 
protein substance surrounding them. This was found when dilute carbol- 
fuchsin was used in combination with light green. Tissue which had 
been killed in chrom-acetic or formal-alcohol fixatives was dehydrated 
and embedded in the usual manner. Sections were cut between 6 and 
12fx in thickness. The sections were stained for one minute in carbol- 
fuchsin which had been very greatly diluted (1 part by volume of carbol- 
fuchsin to 100 parts of water). Then after very rapid treatment with 
