1900] THE HAUSTORIA OF THE ERYSIPHE 165 
haustorium as the collar so frequently mentioned. De Bary (12 
p. 26) probably saw something of this phenomenon in the form 
of the knob-shaped papillz in the cells as mentioned above. 
He saw the structures in cells with thick epidermis and in what 
he thought were no longer growing haustoria. Because of the 
closely connected series of stages which it is possible to secure, 
it is evident that these are not old, disintegrating haustoria. 
Rather they represent the early stages of penetration. At any 
rate, if they represent dead haustoria, they reveal the outlines of 
the cellulose parts none the less, and so tell the story of devel- 
opment. Just what significance there is in this thickening of the 
host wall cannot be conclusively determined without experiment. 
The wall increases not only in bulk, as Ward found, but in quan- 
tity, as has been mentioned. That there’ soon does appear to 
be a disintegration of a portion of this ingrowth, will be seen 
presently. There are not data at hand to determine whether 
the penetrating tube, by means of some chemical substance, 
excites the cell protoplasm to unusual activity in the production 
of cellulose over the region of penetration, or whether micro- 
scopically small needles would cause such a production mechan- 
ically. Again, it is possible that the stimulating agent is the 
atmosphere acting through the wall atthe point made thin by 
the work of the fungus. There would then be the same reason 
for this local thickening as there was for the original production 
of a thick outer wall in the first place. 
Soon the tube, growing with increasing rapidity, overtakes 
the cell in what may be its efforts to protect itself from injury by 
the fungus. Whatever may have been the stimulus which pro- 
duced it, the thickening of the wall ceases after a considerable 
ingrowth, U-shaped in outline, has been formed. The first 
appearance of disintegration is now to be seen in it. The distal 
end no longer has the appearance of cellulose such as is found 
in the cell wall and in the clear basal portion. Some change 
has been wrought in it which has altered its appearance and its 
reaction to stains. This end (jig. ¢) may now be stained 
slightly orange, and it is slightly and minutely granular. No 
