The distribution of mycelium throughout the plant thus com pares sharply with that found in orchids where it is limited to the roots of the adult and the embryonic tissues of the seedling. In Calluna and kindred ericaceous species it is practically co-extensive with the plant tissues. Only the embryo and endosperm of the resting seed are free from contamination; at all other stages of development and throughout its vegetative existence every plant of ling yet examined is a dual organism. In the genus Vaccinium the intimacy of the relation has been carried a stage farther and mycelium is present, not only externally upon the seed-coats, but internally throughout the tissues of the seed. The mode of in fection observed in Ericaceae is unique, for, in all other known cases, the mycelium responsible for forming mycorrhiza enters the root from the soil.
In ling and probably in allied species the formation of func tional mycorrhiza is closely related to soil conditions. It is at a maximum in peaty soils and in cultivated soils with abundant humus, much less conspicuous in sandy soils deficient in organic matter, and may appear to be absent under certain conditions, although roots are never free from fungus infection.
Since seedling development follows and depends upon invasion by fungus mycelium at germination, it is impossible to provide ex perimental evidence bearing on the behaviour of individuals with and without fungus infection.
Woodland soils possess an unusually rich and varied fungus flora among which members of the great group known as Hymen omycetes are conspicuous, as evidenced by the profusion of their spore-bodies—toadstools----produced during suitable seasons. So constant is the association of some of these with certain trees that it has been said that they follow the latter "as the dolphin follows the ship." In order to prove a direct relation between these common woodland fungi and the mycorrhiza of the trees beneath which they grow, it was necessary to isolate the endophytes, learn some thing of their behaviour and compare the growth of young trees with and without mycorrhiza. By such means a number of Hy menomycetes have been identified as the specific root fungi of pine, spruce, larch, birch and other trees.
There is no evidence that seedling development is bound up with fungus infection or that anything in the nature of an obli gate relation exists. As regards the effect of tree mycorrhiza on nutrition, it must be noted that woodland soils possess certain definite characteristics. There is severe competition for water
and salts and they show a notable deficiency of certain plant nutrients, in particular nitrates. On the other hand, they possess large reserves of nitrogen in the form of organic compounds It has been proved experimentally that the possession of mycor rhiza is beneficial to trees, especially in relation to the supply of nitrogenous nutrients. Hence, the supply of nitrates being deficient it may be concluded that it possesses a vital significance for trees and other plants growing in woodland soils. The appli cation of these conclusions to natural conditions raises problems of great importance to practical foresters.
Identity of Mycorrhizal Fungi.—It may be assumed that the fungi responsible for forming mycorrhiza. are widely dis tributed in soil. Their systematic position is known only in those forms which have been isolated and grown in "pure culture" out side the root.
The root fungi of orchids are generally recognized as belonging to a common group, sometimes included in the well-known genus Rhizoctonia, sometimes placed in a new genus Orcheomyces. The mycelium must be locally present about the roots of wild orchids but nothing is known as to its wider distribution.
The mycorrhizal fungi of Calluna and other ericaceous species have been isolated and included in a single species of the genus Phoma, P. radicis. The specialized strains associated with different ericaceous species are distinct and must be locally distributed in heath and moorland soils but there is no certain knowledge of their wider occurrence.
The fungi definitely identified with the mycorrhiza of trees belong to the genera Boletus, Amanita, Cortinarius, Lactarius, Russula and Tricholoma. It is probable that many others will prove to belong to the same category.
Outside these specialized groups, the fungi responsible for mycorrhiza-formation still await identification. The mycelium present in a large number of herbaceous flowering plants, certain ferns and club mosses is uniform in type and produces charac teristic structures : round or pear-shaped swellings or vesicles, variously regarded as organs relating to spore-formation or to the storage of reserves, and intracellular branch systems first described by Gallaud in 1905 under the name of arbuscules. Following upon digestive activity in the host cells the latter disintegrate to struc tureless masses named sporangioles before their true nature is recognized. The fungus responsible for this widespread type of mycorrhiza has hitherto resisted all attempts at isolation and nothing is known with certainty as to its systematic position.