78 A Monograph of the Erysiphaceae 



larger than those upon healthy branches and contained the peri- 

 thecla in all their tissues. For example, a bud well up from the 

 base of a twig might not show any signs of perithecia upon the 

 exterior, but when the large loose scales were removed the bases 

 of the inner ones would expose a dark covering consisting of the 

 mildew perithecia. Longitudinal sections through such buds 

 showed that the living tissue of an ordinary bud was absent, and 

 its space was occupied by an entangled mass of fungous fruit. 

 . . . As fungus parasites thrive upon compounds rich in albu- 

 minous substances, and as the Phytoptus induces a rapid and, 

 therefore, comparatively succulent growth upon a tree that nor- 

 mally has a dense wood, firm bark, and minute, closely protected 

 buds, it may, however, not be so strange that the Sphacrothcca 

 will flourish upon the distortions caused by the mite when it fails 

 to gain a foothold upon a healthy twig. ... It was, moreover, 

 observed that in a cross-section of the stimulated branch there 

 was considerable starch scattered through the bark, while in the 

 healthy and mature twigs, where the buds were normal, smooth- 

 coated, and varnished, there was no starch outside the rino- of 

 firm wood. This starch in the bark may be the secret of the suc- 

 cess of the Sphaerothcca upon the infested branches." 



It is this apparently intimate association of S. pJiytoptopJiila 

 with the galls of the Celtis that suggests the idea of the plant being 

 biologically distinct, and experiments bearing on this point, either 

 the sowing of the conidia of 5. phytoptophila on host-plants of S. 

 hmntili or, conversely, attempts to infect the galls of the Celtis 

 with the conidia of 5. luunuli are very desirable. 



Kellerman and Swingle (198) mention that they were unable 

 to find any haustoria in living specimens, but consider that this 

 may have been due to the distorted nature of the epidermal cells. 



Other cases of the occurrence of species of the ErysipJiaceac 

 on galls caused by mites are known. The form of Micro- 

 sphaera ^////described as '' M. erincophUa"' occurs on "erineum" 

 galls on leaves of Fagus ferniginca ; M. abii also occurs 



under the same conditions on leaves of Cephalanthns occi- 

 dentalis. In both these cases, the presence of the galls seems 

 to cause the appendages of the fungus to become colored, 

 but there is no reason for considering the form a distinct species, 



