46 JOURNAL OF THE ROYAL HORTICULTURAL SOCIETY. 



infection the mycelium of the fungus is safely lodged in the tissues of 

 the host-plant, and no means are known, nor at all like y to be discovered, 

 of killing such mycelium without at the same time killing the portion of 

 the plant containing it. 



The important point to remember in attempting preventive measures, 

 is the fact that the conidia cannot directly infect a Snowdrop leaf, but 

 that inoculation can only be effected by mycelium that has been growing 

 for some time in the soil. Such mycelium originates from sclerotia that 

 have passed the winter in the soil ; consequently, if the soil overlaying 

 Snowdrops is removed during the winter, and fresh uninfected soil added, 

 the danger of infection is reduced. 



If the newly-added soil is mixed with lime — assuming this to be 

 practicable — the mycelium originating from sclerotia attached to bulbs 

 wwld be checked. 



It is the manure and humus in the soil that alone furnishes the 

 mvcelium ^^'ith food and enables it to become sufficiently vigorous to infest 

 the Snowdrop leaves and bulbs. An ideal soil, in w^hich the mycelium 

 could not grow, should contain no organic matter, the necessary plant- 

 food being furnished by artificial fertilisers. Such a condition may not 

 be quite practical ; nevertheless, the ideal should be aimed at as far as 

 possible. Do not use manure, and change the surface soil during the 

 vdnter. 



Description of Figure 4. 



1. A young Snowdrop badly diseased; natural size. 



2. Fruiting branch of the Botrijtis ; x 350. 



3. A Snowdrop bulb with sclerotia ; natural size. 



4. Isolated sclerotia ; natural size. 



5. A sclerotium bearing a crop of Botnjtis ; x 10. 



6. Botrytis conidia germinating ; x 400. 



7. One of the organs of attachment formed on the mycelium of 

 Botnjtis ; x 400. 



8. Chains of colourless conidia-like bodies produced at the tips of 

 special short branches of the mycelium of Botrytis ; x 400. 



