probasidium is short and usually continuous—although it sometimes becomes two-celled—and on this structure is produced a terminal conidium which on germination (in nutrient solution) again produces a probasidium, which in turn gives rise to a conidium; in U. Tritici and U. striaeformis the probasidium gives rise directly to an infection hypha, conidia being absent.

1. Ustilago Avenae Jensen. (Text-fig. 2, and Plate 41, fig. 2.)

Jens., Charb. Cereales, p. 4, 1889. Gramineae.

U. perennans Rostr., Overs. K. Danske Vid. Selsk. Forh. 1890, p. 15, 1890. Cintractia Avenae Ell. et Tr., Jour. Myc., vol. 6, p. 77, 1890.

Sori in spikelets, seldom in the leaves, usually completely destroying the floral parts, forming a pulverulent, olivaceous spore-mass, which eventually falls away, leaving only the bare axis of the inflorescence.

Spores globose or subglobose, 5–9 mmm. diam.; epispore distinctly but finely verruculose, pallid olive, more lightly coloured on one side, 0.5-0.75 mmm. thick.

Host: Arrhenatherum elatius (L.) Beauv. In inflorescences. Herb. No. 1247. Plimmerton (Wellington), sea-coast, E. Bruce Levy! 22 Dec., 1920. E. H. Atkinson! 1 Dec., 1921. Ashburton (Canterbury), 70 m., H. H. Allan! 15 Dec., 1921. Crookston; Dunrobin (Otago), J. C. Neill! G. H. C. 7 Feb., 1924.^†

Distribution: World-wide.

The form on Arrhenatherum elatius was separated as a distinct species on account of the presence of perennating mycelium in the perennial parts of the host; it is identical morphologically with the form on the oat, so that it cannot be considered as more than a biologic race.

Germination.—In water the spores commence to germinate in from twelve hours (fresh material) to three days (old material). A long and slender probasidium (occasionally two) is produced; into this the protoplasm of the spore penetrates, and after the probasidium has attained a length several times that of the diameter of the spore several (3–5) transverse septa appear. On or near these septa several elliptical, hyaline, minute, continuous conidia appear. When detached these may in turn germinate and produce long and narrow infection hyphae. In nutrient solution, according to Brefeld (1883), the conidia do not produce hyphae, but produce secondary or tertiary conidia by a process of budding.

Infection.—This was first studied by Brefeld, who found infection occurred only in the seedling stage of the host. He sprayed budding conidia on to the seedling leaves at different stages of growth; those plants which had attained a development of the leaf beyond the sheathing-leaf were found to be immune. The conidia germinated and produced an infection hypha which penetrated the cuticle and entered the parenchyma, where it produced numerous colourless branched hyphae. As the plant increased in size he found the hyphae increasingly difficult to trace, until at maturity he was able to locate them only in the nodes, where they appeared to be fragmentary and much broken up. Apparently only those hyphae situated in or near the growing-point are able to infect the ovules in the developing inflorescence.

Thus the fungus is transmitted by spores lodged on the seed, between the glume and the pale. When the seed is sown the spores germinate and infect the emerging cotyledom, as described above.

Von Liebenberg (1879) has shown that the spores, if kept under dry conditions, may remain viable for seven years; so that seed stored for several years, if sown without preventive treatment, may bear spores capable of infecting the young plants as they emerge. A second method of infection is known—that of infection from spores remaining in the soil from a previously infected crop. This source is scarcely likely to be troublesome here, for, judging from the readiness with which spores germinate in water, it is probable that with our rainfall such spores would have germinated long ere the seed-bed was prepared for a second crop.

Zade (1922) found that if spores were applied to the inflorescences at the time of flowering, all that fell on the stigmas germinated, producing long probasidia, which in turn produced abundant conidia. When the conidia germinated they produced infection hyphae which gave rise to a mycelium in the peripheral parenchyma of the glumes, the embryo remaining unaffected. He believes this mycelium, and the secondary conidia produced from the hyphae on the inner surface of the glumes, to form the most important source of infection. If this regularly occurs, then dipping of the seed in some fungicide prior to sowing would be useless as a controllant, for the solution used would be unable to penetrate into the glumes to destroy the perennating mycelium therein. In such a case the modified hot-water treatment alone would be of use (see p. 400).

2. Ustilago levis Magnus. (Text-fig. 1, and Plate 41, fig. 1.)

Magn. Abh. Bot. Ver. Prov. Brand., vol. 37, p. 69, 1896.

U. Avenae var. levis Kell. et Sw., Ann. Rep. Kansas Exp. Stn., 2nd Rep., p. 259, 1890.

Sori in spikelets, destroying the inner parts, partially concealed within the glumes, forming a semi-compact, dark-brown, almost black spore-mass.

Spores subglobose, 5–9 mmm.; epispore smooth, dark chestnut-brown, more lightly coloured on one side, 0.75 mmm. thick.

Host: Avena sativa L. In inflorescences. Herb. Nos. 195, 304. Weraroa (Wellington), 80 m., E. Bruce Levy! 12 Jan., 1920. Ruakura (Auckland), 120 m., A. H. Cockayne! 27 Jan., 1921. Lincoln (Canterbury), 30 m., F. E. Ward! 18 Feb., 1921.

Distribution: Probably world-wide, but recorded only from North America and Europe.

This species is abundant throughout New Zealand. It is separated from the preceding on account of the smooth epispores of the spores. The sori also differ somewhat in their being more compact.

Germination, life-history, and control are the same as in U. Avenae.

3. Ustilago Jensenii Rostrup. (Text-fig. 5, and Plate 42, fig. 2.)

Rostr., Overs. K. Danske Vid. Selsk. Forh. 1890, p. 12, 1890.

Uredo Hordei var. tecta Jens., Charb. Cereales, p. 4, 1889. Ustilago Hordei Kell. et Sw., Ann. Rep. Kansas Agr. Exp. Stn., 2nd Rep., p. 268, 1890.

Sori in spikelets, forming a semi-compact black mass, long covered by the transparent basal parts of the glumes.

Spores globose to shortly elliptical, 7–11 × 5–8 mmm.; epispore smooth, olivaceous or sepia-brown, more lightly coloured on one side, 0.75–1 mmm. thick.

Host: Hordeum vulgare L. In inflorescences. Herb. Nos. 298, 1253. Ruakura (Auckland), 120 m., A. H. Cockayne! 25 Jan., 1921. Blenheim (Marlborough), 50 m., J. Scott! 28 Feb., 1921. Winton (Otago), J. C. Neill! 14 Feb., 1924.

Distribution: Australia; North America; Europe.

This is known as the “covered smut of barley,” on account of the manner in which the sori are covered by the basal portions of the glumes. It is common on this host throughout New Zealand.

This species, together with U. Avenae, U. levis, and U. Tritici, were at one time included under the collective species U. segetum Dittm. Jensen (1889A), as the result of numerous infection experiments, split U. segetum into three races, as follows: (1) U. segetum var. Avenae; (2) U. segetum var. Tritici; (3) U. segetum var. Hordei. The variety Hordei he further divided into the two forms: U. segetum var. Hordei forma nuda; U. segetum var. Hordei forma tecta.

Brefeld (1888) considered the forms on barley and on wheat to be distinct, for he found the probasidium did not produce conidia, but gave rise directly to an infection hypha. He found also that these races would not infect oats; consequently he considered them to be distinct from the form on oats, and included them under the name of U. Hordei Bref. In his species are included U. segetum var. Tritici Jens. and U. segetum var. Hordei Jens. (including the two forms nuda and tecta).

Kellerman and Swingle (1890) raised to specific rank the two forms on barley—nuda and tecta—naming them respectively U. nuda (Jens.) Kell. et Sw., and U. Hordei (Pers.) Kell. et Sw. Thus Brefeld's U. Hordei was split by them into three species—(1) U. Tritici (Jens.) Kell. et Sw.; (2) U. Hordei (Pers.) Kell. et Sw.; (3) U. nuda (Jens.) Kell. et Sw.

Now, as the name Hordei was previously used by Brefeld, it cannot again be applied to a species; as U. Hordei of Kellerman and Swingle was named U. Jensenii by Rostrup the same year, this name should be used for this species. The matter is further complicated in that although the forms tecta and nuda are morphologically separable, and may therefore be considered as distinct species, the form nuda (U. nuda Kell. et Sw.) is identical morphologically with U. Tritici Jens., for both possess pulverulent sori, echinulate spores of the same size and colour, and the same method of germination. After extended and critical examination of these two so-called species I am unable to indicate a single morphological character by which they may be separated. True, in certain collections the spore-mass is slightly darker in colour on wheat than on barley, but this character is not distinct in all collections.

I am of the opinion, therefore, that U. Tritici Jens. and U. nuda Kell. et Sw. must be reunited under one name. As to the selection of this name: U. segetum cannot, of course, be used, nor can U. Hordei Bref., for this included the recognized species U. Jensenii Rostr. and the one under discussion. U. Tritici Jens. was by Kellerman and Swingle described on page 262; U. nuda Kell. et Sw. on page 277: since the former has page priority, it is the one to be used.

Clinton (1906) gives the citation of this species as U. Tritici (Pers.) Rostr., but the specific name used by Rostrup is taken from Uredo segetum var. Tritici Persoon, in Tent. Disp. Fung., p. 57, 1797, a publication which antedates the starting-point of modern nomenclature.

The four species under discussion may then be separated by the following characters:—

Germination and infection (of U. Jensenii) are similar to the preceding two species.

4. Ustilago Tritici Jensen. (Text-fig. 3, and Plate 42, fig. 1.)

Jens.: Kell. et Sw., in Ann Rep. Kansas Agr. Exp. Stn., 2nd Rep., p. 622, 1890.

Ustilago segetum var. Tritici Jens., Om. Korns. Brandp., p. 61, 1888. U. segetum var. Hordei forma nuda Jens., Jour. Roy. Agr. Soc., vol. 24, p. 4, 1889. U. Hordei var. nuda Jens., Charb. Cereales, p. 4, 1889. U. nuda (Jens.) Kell. et Sw., Ann. Rep. Kansas Exp. Stn., 2nd Rep., p. 277, 1890. U. Tritici (Pers.) Rostr., Overs. K. Danske Vid. Selsk. Forh. 1890, p. 15, 1890. U. Hordei Rostr., l.c., p. 10. U. Tritici forma foliicola P. Henn., Zeitschr. Pflanzenkr., vol. 4, p. 139, 1894. Ustilagidium Hordei Herzb., in Zopf, Beitr. Phys. Morph. Orq., vol. 5, p. 7, 1895. U. Tritici Herzb., l.c.

Sori in spikelets, destroying ovaries and glumes, forming olivaceous or almost black spore-masses, finally falling away and leaving the bare axis of the inflorescence.

Spores globose to shortly elliptical, 5–8 × 4–5 mmm.; epispore minutely but distinctly verruculose, pallid to dark olive, with a lighter-coloured zone on one side, 0.5-0.75 mmm. thick.

Hosts:—

Triticum vulgare Vill. In inflorescences. Herb. No. 407. Lincoln, Canterbury, 30 m., F. E. Ward! 5 March, 1921.

Hordeum vulgare L. In inflorescences. Herb. No. 46. Weraroa, (Wellington), 120 m., G. H. C. 11 Nov., 1919.

Distribution: World-wide.

The form on wheat is commonly known as “naked smut on wheat,” that on barley as “naked smut of barley.” As has been pointed out (under U. Jensenii), the two forms differ only in that each is confined to its host, but as they are identical in morphological characters they must be considered as the same species.

The two species on barley may be separated on account of the following differences:—

Germination.—In water a probasidium is produced which gives rise directly to an infection hypha; this may branch and form a mycelium, but does not at any time produce conidia.

Infection.—With the race on barley it was first suggested by Jensen (1889B), and later independently confirmed by Hecke (1905), that infection occurs through the flower. With the race on wheat Maddox (1897) first demonstrated that infection occurred through the flower; his work was later confirmed by Brefeld (1903). The spores are deposited by wind on the stigmas at the time of pollination, where they germinate and produce a probasidium, which develops immediately into an infection hypha; this penetrates the style and enters the cells of the developing ovary, where a mycelium is formed. As the embryo commences to develop, hyphae pass into it and form a resting mycelium. which remains latent until such time as the grains germinate. When germination commences these hyphae grow out with the cotyledon, keeping pace with the growing-point until the formation of the inflorescence, when they penetrate to the developing ovaries. These they convert—together with the developing ovules—into a mass of mycelium, which later produces the spores.

The significance of this life-history, from the viewpoint of the farmer, is that the resting mycelium is embedded in the tissues of the grain, where it cannot be reached by fungicides; consequently the disease cannot be controlled by the usual steeping methods. Freeman and Johnson (1909) claim to have successfully combated this smut by means of Jensen's modified hot-water treatment (see p. 400).

5. Ustilago striaeformis (Westendorp) Niessl. (Text-fig. 7, and Plate 46, fig. 2.)

Niessl, Hedw., vol. 15, p. 1, 1876.

Uredo striaeformis Westnd., Acad. Roy. Belgique, Bull. 18, ser. 2, p. 406, 1852. Tilletia De Baryana Fisch. v. Waldh., in Rabh. Fungi Eur., No. 1097, 1866. T. Milii Fcl., Symb. Myc., vol. 1, p. 40, 1869. T. striaeformis Oud., Bot. Ztg., vol. 36, p. 440, 1878. T. alopecurivora Ule, Bot. Ver. Prov. Brandenburg, vol. 25, p. 214, 1884. T. Brizae Ule, l.c. Ustilago Poarum McAlp., Proc. Roy. Soc. Vic., n.s., vol. 7, p. 220, 1894. U. washingtoniana Ell. et Ev., Bull. Torrey Cl., vol. 22, p. 57, 1895. Tilletia Airae-caespitosae Lindr., Soc. pro Fauna et Flora Fennica, vol. 26, p. 15, 1904.

Sori rarely in the inflorescences, commonly in the leaves and leaf-sheaths, forming long striae often many centimetres in length, frequently converging to form irregular black areas, at first covered by the epidermis, becoming exposed and pulverulent, finally in old specimens the leaves becoming shredded; in inflorescences the sori are frequently hidden within the glumes, but eventually become pulverulent and partially exposed.

Spores globose to shortly elliptical, often somewhat angular, 10–17 × 8–12 mmm.; epispore moderately and somewhat coarsely verrucose, olive-brown, 1–1.5 mmm. thick.

Hosts:—

Dactylis glomerata L. In leaves and sheaths. Herb. No. 265. Cluny, Turakina (Wellington), 210 m., G. H. C. 2 Jan., 1921.

Holcus lanatus L. In leaves. Herb. No. 1276. Ettrick (Otago), 300 m., R. B. Tennent! 10 Feb., 1921. Karori (Wellington), 200 m., R. Waters! 6 Dec., 1923.

Distribution: World-wide.

This species is readily recognized by the manner in which the infected leaves become shredded, tearing occurring along the linear sori. Infected plants are usually stunted, and not infrequently killed outright.

Germination.—Osner (1916) has well illustrated the method of germination. A short probasidium is produced, and this grows directly into an infection hypha; no conidia are produced.

Infection.—According to Osner, healthy plants become infected through the flowers, as is the case with wheat or barley infected with U. Tritici. Following infection the hyphae remain quiescent in the seed until they germinate, when they grow out with the developing leaves and stems. These hyphae give rise to the linear sori; those that grow with the growing-point of shoots producing inflorescences infect the ovaries as they are formed. The mycelium also hibernates in the roots of perennial plants, and infects developing shoots as they emerge in the spring. Osner points out that the disease may be combated by means of the modified hot-water treatment (p. 400), but this would prevent only young plants (seedlings) from becoming infected, for, since he has shown the mycelium to be perennial in perennial plants, it follows that once such a plant has become infected nothing in the way of remedial treatment can be attempted.

The species is placed by the majority of systematists in the genus Tilletia, largely on account of the work of Fischer von Waldheim (1869), who claimed the method of spore-formation was that of this genus. Osner, however, has shown that spore-formation is intercalary in the spore-forming hyphae, and that this “adds weight to the contention that the organism is a species of Ustilago rather than of Tilletia.” He mentions the appearance of transverse septa in the probasidium as another character in support of placing the species in Ustilago. This, however, is of little significance, since these septa also appear in the probasidia of Tilletia.

6. Ustilago bromivora (Tulasne) Fischer von Waldheim. (Text-fig. 8, and Plate 43, fig. 1.)

F. v. Waldh., Bull. Soc. Nat. Mosc., vol. 40, p. 252, 1867.

Ustilago Carbo var. vulgaris d. bromivora Tul., Ann. Sci. Nat., ser. 3, vol. 7, p. 81, 1847. Cintractia patagonica Cke. et Mass., Grev., vol. 18, p. 34, 1889.

Sori in spikelets, often hidden within the glumes, sometimes destroying them at the base, at first bullate and somewhat compacted, becoming pulverulent, black.

Spores globose to shortly elliptical, often polygonal, 8–11 × 7–10 mmm; epispore minutely but densely verrucose, olivaceous or dark reddish-brown, 1 mmm. thick.

Hosts:—

Bromus hordeaceus L. In panicles. Herb. No. 442. Blenheim (Marlborough), 30 m., F. Sisson! 12 Nov., 1920.

Bromus unioloides H. B. K. Herb. No. 47. Weraroa (Wellington), 120 m., G. H. C. 11 Dec., 1919. Omahu, Thames Valley (Auckland), W. G. Goodwin! 22 Nov., 1921. Blenheim (Marlborough), 30 m., R. Waters! 27 Nov., 1923. Lincoln (Canterbury), 30 m., F. E. Ward! 3 Dec., 1923.

Distribution: World-wide.

The species is fairly abundant on prairie-grass throughout New Zealand. In many cases only occasional spikelets may be infected, but as a rule all on a panicle are attacked.

Germination.—In water the spores readily germinate, producing a small probasidium, on the apex of which is produced a single elliptical conidium. In nutrient solution, according to Brefeld (1883), a two-celled probasidium is formed; this produces conidia, which in turn develop a probasidium again producing conidia.

Infection occurs in a manner similar to U. Avenae. McAlpine (1910) states that treating the seed with formalin or copper-sulphate effectively controls this smut.

7. Ustilago bullata Berkeley. (Text-fig. 6, and Plate 42, fig. 3.) Berk., Fl. N.Z., vol. 2, p. 196, 1855.

Sori in inflorescences, at first enclosed in a lead-coloured membrane, when semi-compact, becoming exposed, when pulverulent, black.

Spores globose to shortly elliptical, 8–12 × 7–8 mmm.; epispore closely and finely verrucose, olivaceous, 1 mmm. thick.

Host: Agropyron scabrum (Lab.) Beauv. In inflorescences. North Island, W. Colenso. 1849. (Type in Herb. Kew.)

Distribution: Australia.

The host is indigenous and widespread; it occurs also in Australia (Cheeseman, 1906, p. 923).

This species was described by Berkeley from material forwarded by Colenso to Kew. I have no New Zealand material in the herbarium, the above description being drawn up from Australian material kindly forwarded by Mr. C. C. Brittlebank, Plant Pathologist, Department of Agriculture, Melbourne.

Germination.—McAlpine (1910, p. 152) has successfully germinated the spores of this species. He found that in water the probasidium became three-celled, the proximal cell eventually developing into a well-developed infection hypha; no conidia were produced, but in nutrient solution a probasidium was produced, which gave rise to numerous conidia.

8. Ustilago comburens Ludwig. (Text-fig. 4, and Plate 41, fig. 3.)

Ludw., Zeitschr. Pflanzenkr., vol. 3, p. 139, 1893.

Ustilago microspora Mass. et Rodw., Kew Bull., p. 160, 1901. U. exigua Syd., Ann. Myc., vol. 1, p. 177, 1903.

Sori in spikelets and stems, at first compact, bullate, and covered with a lead-coloured membrane, later destroying the entire panicle and becoming exposed on the rachis as a dense bronze pulverulent mass, eventually falling away, leaving the naked axis.

Spores globose to shortly elliptical, 3–5 × 2–3 mmm.; epispore smooth, tinted olive, 0.5 mmm. or less in thickness.

Host: Danthonia Buchanani Hook. f. In panicles. Herb. No. 497. Dunstan Mountains (Otago), 350 m., W. D. Reid! 6 Dec., 1921.

Distribution: Australia.

The host is endemic, and is confined to the mountain regions of the South Island (Cheeseman, 1906, p. 891).

The species is characterized by the exceedingly minute size of the spores. It was first recorded on a species of Stipa, but McAlpine (1910, p. 154), who has examined portion of the type material, states that the host in question is a species of Danthonia.

I have been unable to germinate the spores.

9. Ustilago Readeri Sydow. (Text-figs. 9, 27, and Plate 43, figs. 2, 3.)

Syd. in letter; McAlp., Smuts. Austr., p. 159, 1910.

Ustilago Agropyri McAlp., Ag. Gaz. N.S.W., vol. 7, p. 154, 1896.

Sori in spikelets, stems, and leaves, commonly destroying the entire inflorescence. On stems and leaves forming conspicuous striae; in inflorescences, at first enclosed within the leaf-sheath, becoming exposed and appearing in the form of semi-compact masses covered with the remnants of the glumes, when greyish in colour, at length naked, pulverulent, black, finally falling away leaving the bare axis.