170 BOTANICAL GAZE TTE [MARCH 
not ruptured by the ingrowth of cellulose. Rather the ingrowth 
is caused by the activity of the membrane, the area of which 
enlarges with the thickening of the wall. The plasmic mem- 
brane is still recognizable when the haustorium begins to enlarge 
at its distal end. It still maintains its osmotic properties at the 
stage represented by fig. 7. In most cases it remains bounding 
the masses of disintegrated cellulose which constitute the contents 
of the sheath. Fig. 7 represents the membrane stretched and 
firm from the osmotic forces at work in the nutrition of the 
fungus. At a later stage it usually suffers injury and some- 
times dissolution. 
This fungus is not capable of producing the extraordinary . 
dissolution of cellulose which Ward found in the case of Botry- 
tis. Its supply of enzymic material seems to be limited. The 
sheaths are, therefore, usually present, though on rare occasions 
they are not, as in jigs. 8 and 12. Certain haustoria of Uncinula 
Salicis do not possess sheaths and, as will be seen, this species 
has a greater digestive capacity.  Nordhausen (23, p. 23) found 
in infecting leaves with spores of Botrytis that heavy dews so 
weakened the enzyme of the fungus that penetration was impos- 
sible. It seems probable that, when the cell-sap of Geranium 
begins to be absorbed by the young haustoria of Erysiphe, the 
enzymic material is weakened or largely reabsorbed. Thus it 
can be seen how the sheaths with their bounding membranes are 
possible and how, as De Bary observed long ago, the host-cells 
may escape with so -little injury. From the time when the 
haustoria begin to absorb actively, the further digestion of the 
cellulose papillz largely ceases, and though their bounding i 
branes collapse, they are usually not completely digested, being 
Protected by the cell-sap. In fig. 8 the sheath is all digested 
down to the collar. In Jig. 19 the haustoria succeeded in pene 
trating the wall of the hair, of which the cell forms a part, before 
the plasmic membrane could build up papille to impede the 
Progress of the tubes. The digestive capacity of the haustor# 
is not entirely constant, so that we find sheaths in various stage 
of disorganization. 
