Various aspects of smut infection in grasses, mostly subtropical perennials, have been elucidated. The research has included field and glasshouse observations of variation in smut expression in individual tussocks, investigation of host-parasite relations, especially the distribution of hyphae in the hosts, soral development and sporogenesis, and ultrastructural studies of intracellular hyphae.
The tillering behaviour of Heteropogon contortus infected by Sorosporium caledonicum has been studied during three consecutive growing seasons, the growing season being that period between September each year and the following May when the plants are growing actively. It has been established that smut-infected tussocks which produce smutted tillers in the early part of the season may bear both smutted and flowering tillers in the latter part of the season. The flowering time for smut-infected tussocks is the same as that for smut-free tussocks. The flowering tiller of a smut-infected tussock, in the following season, produces from its base further tillers. The latter may bear either smut sori or normal inflorescences.
A study of hyphal distribution in tiller bases of H. contortus has shown that hyphae may move into axillary buds from which tillers develop at various stages of bud and tiller development. Hyphal distribution in the tiller base determines whether an axillary bud receives mycelium; and the extent to which hyphae grow through a tiller's tissues determines whether a tiller bears a sorus or an inflorescence. Soral formation occurs only if hyphae are close to the growing point at the time of floral initiation.
By a study of 13 smuts on 12 different grasses, and of a smut of Polygonum, the characteristics of sori were determined. The degree of destruction of host tissues and the complexity of structure of the sori resulting from proliferation of cells of the host and the fungus are the criteria for recognizing three types of sorus. The origin and development of columellae and peridial structures and processes of sporogenesis have been elucidated. Electron microscopy has been used for detailed observation of the early stages of spore formation. By the same technique it has been shown that the intracellular hyphae of smut fungi resemble haustoria in many respects, and that possibly there is no essential difference between them.
The taxonomic implications of data acquired by studies of the ontogeny and development of the various components of sori have been considered. Evidence to reinforce earlier opinions that Sphacelotheca is not a valid genus has been presented. A new genus Tectosporium, with Tectosporium andropogonis as the type species, has been proposed for smuts with sori bounded by a peridium composed of both host cells and hyphae, and having sporogenous hyphae in pockets separated by partitions of non-sporogenous hyphae, the latter at soral maturity providing groups or chains of partitioning cells that lie among the teliospores.