THE HONEY FUNGUS 163 



only one tree, a Corsican pine, has so far been attacked. 

 (There is no plot of Scots pine.) Not a single larch, deodar,i 

 or Norway spruce has been affected, though from other 

 woods I have obtained specimens of each of these species 

 killed by the fungus. If perfectly healthy roots were liable 

 to be penetrated by the rhizomorphs, I do not think that such 

 immmiity could have been expected. 



In a larch tree in a neighbouring plantation , about eighteen 

 years old, which showed only the early stages of attack, I 

 found that the rhizomorphs had entered through a wound 

 in the bark at the base of the trunk. In this case the roots 

 and rootstock presented more than a square foot of un- 

 damaged bark, and yet the fungus found out the square 

 inch of wound. Further instances might be quoted from 

 other observations, but enough has been said to show the 

 improbability of infection through healthy bark. 



From Brefeld's experiment we know that rhizomorphs 

 can enter dead roots, and it is jDrobable that this faculty is 

 commonly employed by the fungus. The importance of 

 dead roots has been fully discussed in the chapter on Fomes 

 annosus, and need not be reiterated ; and though this 

 method of attack by Armillaria raises some further questions, 

 these have not yet been investigated. Infection through 

 dead roots may account for a phenomenon that is frequently 

 observed — I refer to the case of larch woods which remain 

 practically free from infection by Armillaria until they 

 reach an age of more than fifteen years, after which many 

 trees succumb within a short time. The presence of the 

 fungus may have been detected many years previously, 

 but it had to wait until the death of some of the tree 

 roots afforded a means of entrance into the living trees. 



3Ieans of prevention. The forester is too often taken 



unawares by the honey fungus, but generally a surprise is 



due to lack of previous observation. When oak or beech 



woods are cut do^vn the fungus must be expected, and any 



coniferous plantations on a site which has previously been 



occupied by such trees are especially liable to attack. 



^ One deodar was killed this year (1919). 

 M 2 



