S. p. Wiltshire 187 



ragged membrane over the scar. Pustules of conidia are sometimes found 

 on the primary scars in January and February and these show the extra- 

 ordinary vigour with which the fungus flourishes during the dormant 

 period of the host. Plate III, Fig. 2 shows young infections on the 

 Medaille d'Or x Kingston Black seedhngs, which only become infected 

 in the late spring and the cankers therefore develop slowly. Typical leaf 

 scar infections as they appear in summer are shown in Plate III, Fig. 3. 

 The development of the canker after bud infection may be so rapid 

 that the whole of the shoot above it is killed almost immediately. On 

 the other hand, development may be slow and the canker may take 

 three or more years before it completely encircles the stem. It is possible 

 to find many cases of cankers on old wood which can well be attributed 

 to leaf scar infection (Plate III, Fig. 4). 



MICROSCOPIC DETAILS. 



Observation of the development of the canker showed that in many 

 cases at least, the infection started from the leaf scar and not from the 

 bud itself. The leaf scars were therefore examined minutely. The dead 

 tissues of the leaf scar were invariably found to be infested by fungi, one 

 interesting fungus not yet identified being almost universally present. 

 It has been found on the leaf scars within a week of defoliation and very 

 rapidly gains a foot-hold on the scar tissue. It has dark coloured my- 

 celium and is associated with a pycnidium containing dark, oUve green, 

 muriform pycnospores (Text-fig. 2). The pycnidia are sometimes immersed 

 in the leaf scar tissue or they may occur outside seated on a small mass 

 of hyphae. In addition to the pycnidia, dark coloured conidiophores 

 are commonly found bearing spores similar to Fumago vagans, but 

 whether these fructifications belong to the same fungus as produces 

 pycnidia is unknown. The fungus has been noted upon specimens 

 from Cambridge and Worcestershire and is probably of common occur- 

 rence. Whether this fungus plays any part in bringing about infection 

 is not known, but such a possibiUty cannot be overlooked and therefore 

 its occurrence is recorded here. 



In the infection of leaf scars in the early autumn, no case of infection 

 was discovered whilst the leaf remained attached to the tree, although 

 numerous instances were found but shortly after defoHation. Inoculation 

 experiments through the scars of freshly removed leaves during June 

 1919 were completely unsuccessful, although two inoculations of wounds 

 with the same inoculant produced typical cankers. At the normal fall 

 of the leaf in the autumn, however, the host is not in as active state of 



