cellular, and pervades all parts of the affected plants, especially the cortical parts of the stem, and also to some extent the pith. It also occurs in the upper part of the root-stock. It is hyaline, about 4'5μ wide, richly branched, and septate. Its contents are colourless and vacuolated. The spores are formed inside the mycelial hyphæ, where it becomes coarsely granular, much after the manner of Entyloma.* Woronin † finds that the mycelium is abundantly provided with very marked haustoria, which enter the cells. These botryform prolongations enter and occupy a third or a half of their interior. Germination takes place in autumn. Specimens gathered in June germinated in October and November. The epispore splits, and the endospore grows out as a blunt cylindrical promycelium. At its extremity it emits a cluster of from four to seven apical branches. The outgrowth of the endospore is at first often in the form of two equal branches, one of which develops into the promycelium, while the other ceases to grow and has become emptied of its protoplasm, which passes into the developed branch. Towards the upper half of the promycelium a septum appears, cutting off the protoplasm above from the empty tube below; but true spore-formation was not observed.

Tubercinia. The germination of the spores of T. trientalis has been worked out by Woronin. Teleutospores collected at the end of September and the beginning of October were found often to have already germinated upon the plant. Placed in a damp atmosphere, they germinated freely after the manner of Tilletia, each spore producing a promycelium surmounted by a cluster of spores. All the teleutospores of one spore-ball do not germinate at the * De Bary, loc. cit.

† Woronin, loc. cit., pp. 27, 28, t. iv. figs. 27-35.

Woronin, "De Bary und Woronin Beitrage," 5 reihe (1882), pp. 4-16, t. i., ii., iii. figs. 1-12.

same time. The promycelium emerges through a small round opening in the epispore, its length corresponding to the size of the spore. On the upper, free, blunt end of the promycelium from four to eight protuberances appear, which elongate themselves into branches and become the cylindrico-fusiform promycelial spores. After all the protoplasm from the interior of the teleutospores has been passed into the upper end of the promycelium and into the developing promycelial spores, a septum is formed close to its upper end. If the promycelium happen to be a very long one, two or more septa occur. The promycelial spores, while still attached to the promycelium, become united in pairs by a bridge-like connection. This conjugation takes place at the bases of the promycelial spores, and but rarely at their summits. One of the conjugated spores then buds out a secondary spore, which in its turn sometimes produces a tertiary; sometimes all these may be observed in a chain. If there be an odd spore on the promycelium which has not conjugated, it does not bud.

Conidia.*-These are produced from a mycelial mesh that exists for the most part just beneath the epidermal structures, and is provided with very numerous botryform haustoriæ (Plate VI. Fig. 1). The conidiophores emerge through the stomata, or between the epidermal cells (Plate VIII. Fig. 1). The conidia are borne almost horizontally; they are from II to 15μ long, and consist of subpyriform cells attached by their larger end. A thin hyaline membrane encloses the granular protoplasm, in which a small vacuole may be observed. When placed in a damp atmosphere the vacuole enlarges and a germ-tube is produced, generally from the larger end of the conidium ; into this germ-tube the protoplasmic contents of the conidium are received and passed onwards as it elongates

* Woronin, loc. cit.

H

(Plate VIII. Fig. 2). If the conidium germinate upon a leaf, the germ-tube squeezes its point between the two epidermal cells (Fig. 3), and soon produces in the leaf a mycelium with haustoria. In from twelve to twenty days after infection this mycelium produces the black teleutospores, but not the conidia.

The life-history of this species is peculiar: the teleutospores germinating in autumn produce promycelial spores, which, entering the young subterranean shoots of the hostplant, develop a mycelium, which remains quiescent during the winter, and in the spring produces, first the conidia on the leaves, and afterwards teleutospores mostly in the stem.

The entrance of the germ-tube in this species is (as already stated above) between the epidermal cells. It grows downwards in the partition wall, splitting it into two lamina, and so makes its way through the epidermis.

CHAPTER XI.

INFECTION OF THE HOST-PLANTS BY THE USTILAGINEÆ.

THE manner in which the Ustilagineæ gain admission into their respective host-plants has been studied very carefully by many botanists, but is not yet fully understood. With those species which affect the flowering parts of annual graminaceous plants, such as Tilletia tritici and U. segetum on wheat, it is noteworthy that not only are all the blossoms or fruits upon an ear affected, but also all the ears which arise from one plant. It is very exceptional ever to find one sound ear upon a plant of which the others are diseased; in like manner, it is very unusual to find a sound kernel upon an ear in which the other kernels are affected. Coupled with the fact that in diseased plants the mycelium of the fungus can be found in all parts of the axis, it is obvious that the parasite gained admission into the plant at an early stage of its growth. Kühn* specially investigated this point with T. tritici, and found, in very young wheat seedlings, that the mycelium was present in them. Hoffmann† came to the conclusion that the spores entered between the split in the young sheath and the rootlet. He also figures the spores forming a mycelium which enters the stomata of the young * Kühn, loc. cit., pp. 48, 49.

Hoffman, "Flugbrand," pp. 202-206, t. xiv. figs. 14-18.,,

plant, but this is probably incorrect. To Wolff,* however, we owe the first accurate explanation of this process. He investigated it with U. segetum and maydis, Urocystis occulta and T. tritici. The outcome of his observations is that the germ-tube of the promycelial spores of the species is capable of piercing the embryonic plant at any time before the primary enveloping sheath of the young plant is ruptured. The germ-tubes of T. tritici squeeze their points through the epidermal cells of the young plant, at first piercing through the outer epidermis of the primary sheath ; they then grow through the cells of the sheath itself, then through the inner epidermal cells of the sheath, across the interspace to the outer epidermal cells of the embryo, and so into the embryo itself (Plate VI. Fig. 2). With certain species the entering germ-tube acquires for itself an investing sheath from the cells through which it passes (Plate VI. Fig. 3)-a sort of invagination of the outer wall of the outer epidermal cell, which is continued over the young mycelium as it grows through one cell after another. With Urocystis occulta the investing sheath exists only where the mycelium. passes through the first epidermal cell. Kühn† subsequently repeated and confirmed Wolff's observations as far as they went, but he also found that the germ-tubes could enter, not only through the primary sheath-leaf, but also into the true root node at the base of the sheath, and, in point of fact, into almost any part of the embryo. While it has long been known that, by merely dusting wheat with the teleutospores of T. tritici and planting it, it became affected with bunt, yet with U. segetum such dusting rarely, if ever, succeeds in producing the disease. Hoffmann was able to produce only a few smutted plants in many hundred 'Roggenstengelbrand," Bot. Zeitung (1873), t. viii.; “Der Brand des Getreides" (1874), pp. 18-24, t. iii., iv.

* Wolf,

↑ Kühn, Bot. Zeitung (1874), PP. 121-124; Fahlingsladw. Zeitung (1879)