W. SOMERVILLE 335 



it is a great mistake to purchase plants by size rather than by age. It is 

 also an advantage to stock an area with many more trees than can find 

 room in even early middle life, for thereby one has the opportunity of 

 selecting robust individuals to stand for the final crop, and in this way 

 the financial returns are much improved. Even in respect of improve- 

 ment by accidental or deliberate breeding, there is the great difference 

 between agriculture and forestry, in that whereas agricultural plants 

 are, with hardly an exception, hermaphrodites, trees are with few 

 exceptions unisexual. Self-fertilisation, with the opportunity it gives of 

 perpetuating pure strains, while the rule in agricultural crops, is com- 

 paratively rare in trees. It is impossible in Salix and Populus, and more 

 or less of an accident in the case of such important genera as Quercus, 

 Pinus, Picea, Larix, Abies, and many others. 



While impressed with the special difficulties of Research in forest 

 mycology I am far from suggesting that there are not many hopeful 

 lines that may be pursued. Take, for instance, that most serious pest 

 Peridermium Strobi. I will not dwell upon its easily avoidable intro- 

 duction to North America, where it is threatening the existence of the 

 most important single lumber species of the North American continent, 

 the White or Weymouth Pine. But even in this country its biology is 

 something of a mystery. Its Cronartium stage on Ribes is supposed to 

 be dependent on fresh annual infection with aecidiospores produced by 

 the Weymouth Pine, and yet we are assured by competent observers 

 that in the East of England this Rust occurs abundantly on currant 

 bushes many miles removed from the opportunity of aecidial infection. 

 The question one would like answered is this, is sach infection actually 

 produced by aecidiospores, borne by the wind or in some other way, or 

 is it due to uredospores hibernating in fallen leaves ? I know of no analogy 

 in support of the latter suggestion, for although Rust of wheat can exist 

 away from the Barberry, it seems to be proved that in this case the uredo 

 stage survives the period between one crop and the next on what the 

 Americans call "volunteer" plants, which are usually not difficult to 

 find in neglected corners, on manure heaps, and in other places. Then, 

 again, in America it has been observed that outbreaks of the aecidial 

 stage occur sporadically many miles away from the infected Ribes, and 

 it is suggested that the sporidia or, it may be, the teleutospores are 

 borne on the bodies of caterpillars of the Gypsy Moth, which are said to 

 be wind-borne over long distances. But the difficulty is to reconcile this 

 statement with the fact that in Bagley Wood near Oxford there is a 

 plantation of Weymouth pines whose North margin comes close up to 



