spores in the spore-bed.

*

Whether this be the simple passive process we have hitherto regarded it or not is undetermined. It is quite possible that it may be aided by a projective force, as Brefeld has shown in the Agaricini. In them the basidia become by degrees fuller of protoplasm, until a certain state in the tension of the elastic walls takes place. When this has attained its full extent, the upper part of the basidium, immediately below the spore, gives way and becomes split off all round, the spore being projected from its sterigma at the point of its attachment. A similar process occurs in Pilobolus, only in this case the rupture is lower down, and not at the exact point of junction of the spore with its basidium.

As soon as each uredospore is ripe it is capable of germinating, and when placed in a sufficiently damp environment it does so in a few hours. This process is accomplished in the same manner as has been already described under the æcidiospores. It consists in the protrusion of a germ-tube through one or more of the germ-pores, which branches, elongates, circumnutates, and receives the protoplasm from the interior of the spore, and passes it onwards to its peripheral extremity (Plate III. Figs. 11, 12, 13). In the same way its extremity, or the extremity of one of its branches, enters into one of the stomata of the host-plant, and in its tissues develops a fresh mycelium (Plate III. Fig. 15).

Under certain circumstances, when the germ-tube cannot enter a stoma, instead of growing in the mode described, it dilates in a bulbous manner† at its extremity into a spherical dilatation, into which the orange granules accumulate; or it may give off one or more lateral out

Brefeld, "Schimmelpilze."

Plowright, "Germination of Uredines;" "Grevillea," vol. ix. pl. 159, figs. 10, 11, 12.

growths, which become globose and filled with endochrome (reserve spores) (Plate III. Fig. 14). These abnormal phenomena show how nearly allied the different spore-forms of the Uredineæ are; and are especially interesting from an evolutionary point of view. They show that although the uredospore has attained a definite mode of germination proper to itself, yet it has not entirely lost its capability. of reverting to, and simulating the mode of growth of other spore-forms."

*

In certain species (Lecythea) are found bodies known as paraphyses, or cystidea. They are in most cases arranged around the circumference of each spore-bed, arching over it and arising from basidia in the same manner as the spores themselves. The paraphyses contain no coloured endochrome. In shape they may be globose, pyriform, subcylindrical, or capitate (Plate III. Figs. 16, 17). Although they cannot be regarded as undeveloped spores, inasmuch as they are pretty much confined to particular species, and do not occur indiscriminately with all, yet from their mode of origin their affinity is clearly with the pseudoperidial cells of the acidiospores, and to a certain extent they are protective organs in the same way as the latter. From my observations it appears that the presence of paraphyses with certain species greatly depends upon some special condition of the fungus, as they may be absent, or nearly absent, according to special circumstances. I find them constantly present with the uredospores of P. perplexans, when these have arisen not directly, but rather at a considerable distance, from the acidiospores. On the other hand, when the uredo arises directly from the acidiospore, they are hardly present at all; this looks very much as if they were an indication of exhaustion of vital energy

* Tulasne figures a similar condition in the germination of the uredospores of Cronartium (" 2o Mémoire," tab. xi. figs. 3, 8, 9).

D

on the part of the fungus, which was combated by protective efforts on the part of the parasite in conserving those spores which it does produce, but when full of vigour and fresh from the acidiospore it is less careful of its spores. When it begins to feel the effect of exhaustion, and is unable to develop such energetic spores, it takes more care of those which are produced.

The same is true of the quantity of uredospores themselves. When the uredospores are produced directly from the acidiospores they are much less copious than when they originate from other uredospores; especially is this so when they have arisen from a long series of uredospores. The converse is observable in such species as Puccinia tragopogonis, where the teleutospore occurs often on the same leaf as the ecidiospore, and the uredospores are very few indeed. We have striking illustrations of the contrary condition. with many heterocious Uredines. In Australia, where the barberry is not an indigenous plant, and occurs only in gardens and shrubberies, the agriculturists complain not of 7 mildew (P. graminis) as destroying their wheat crops, but of rust. Some years ago I received specimens of the affected wheat plants from New South Wales, Queensland, and South Australia, all of which showed a profuse development of uredospores in proportion to the teleutospores, quite out of all parallel to that which obtains in England. Mr. C.J Arthur informs me that this is equally true of those districts in the United States from which barberries are absent. The same occurs in this country with P. rubigovera, the uredospores of which are extremely abundant on our wheat crops in spring; but the autumnal Æcidium is a very infrequent fungus, partly because the host-plants are none of them very abundant, and partly because, occurring

* Plowright, "Reproduction of Heterocious Uredines," Jour. Linn. Soc. Botany, vol. xxi. p. 368.

as it does in September and October, the frosts of winter. soon destroy their foliage, so that the uredospores have to reproduce themselves throughout the winter and spring months on the wheat plant. Rostrup* has remarked the same fact with regard to Coleosporium senecionis, that when it occurs in localities from which fir-trees are absent it consists almost wholly of uredospores. Independently of this, however, the uredospores of some species are much more abundant than of others; in P. oblongata, for instance, they occur in great quantities, while in P. hydrocotyles the teleutospores are very few in number, and occur in the same spore-beds as the far more numerous uredospores.

Although uredospores have their vitality so easily destroyed by heat and dryness, they can withstand a considerable amount of cold, for freshly developed spore-beds of P. rubigo-vera can be found almost at any time, on wheat, during the winter months. Of course, it may be that these have been developed by mycelium produced from a spore-infection effected at an earlier date. I found that the uredospores of this species which had been exposed to several nights of frost, when the thermometer registered 23° F. (— 5° C.), germinated with the greatest freedom. Dietel found the uredospores of Phragmidium obtusum, which had been covered by snow from December 18 to January 28, germinated in a day and a half in a warm room.†

p. 9.

*Rostrup, "Heterociske Uredineen" (1884), p. 6.

† P. Dietel, "Beiträge zur Morphol. und Biol. der Uredineen" (1887),

CHAPTER VI.

TELEUTOSPORES.

AT some period in the life of all Uredines* a spore is produced from one of the before-mentioned mycelial developments, to which the appellation teleutospore has been given. In the strict sense of the term teleutospore means the last-formed spore-that is to say, it is formed later in point of time than the acidiospores and uredospores; but while this is true in the majority of cases, yet there are many species of the Uredinea in which no other spore-form than the teleutospore occurs. Still, the name has become so familiarized to us by long usage that it is unadvisable to change it. These spores are formed in a similar manner to the uredospores from a stroma (spore-bed, clinode, hymenium) produced by an aggregation and entanglement of mycelial hyphæ placed just beneath the cuticular structures of the host-plant and parallel to them. Perpendicularly from this stroma are given off erect branches, which, becoming dilated at their free ends, are soon in

* De Bary ("Vergl.," p. 305) points out that Uredo symphyti has no other spore-form than the Uredo, and indicates this species as being a degenerate type, which, having lost its other spore-forms, is capable of existing without them. Without questioning the truth of this statement, one cannot fail to remember that until quite recently the other spore-forms of many Uredines were unrecognized, for instance, Melampsora cerastii ; so it is quite possible that the teleutospores of U. symphyti may exist, although we at present do not know of them.