Anamorph: conidiomata embedded in both sides of the leaf, scattered, inconspicuous, often eaten away probably by the microfauna, minute, circular; smooth, not raising the surface of the substrate, 180-250 µm diam., dark grey to black in the centre, elements of the epidermis remain above the upper with a thick black perimeter line, often with a wall of the conidioma, the centre of the conidioma golden-brown tint to the central area when tanned often coinciding with a single stoma of the leaf. In vertical section: conidiomata composed of an upper and a lower wall, the upper wall being located below the cuticle and the top half of the broken epidermal cells in some of which tanned contents remain, the upper wall being limited in extent to the immediate vicinity of the central stoma, filling the area below it with thin-walled, dark brown cells, 2-3 µm diam., forming a textura angularis sometimes several cells deep; the lower wall extending across the whole conidioma, adjacent at its edge to the cuticle (hence causing the appearance of the black perimeter line), more deeply embedded in the centre, depressing the degraded mesophyll, being composed of thinwalled, dark brown cells forming a layer 1-2 cells thick, above which is a further layer of hyaline, thinwalled cells 2-3 µm diam., several cells thick, forming a rather indeterminate textura angularis from which the conidiogenous cells and trichogynes arise. In horizontal section: the upper wall having a poorly defined textura angularis, often with a minute ostiole 6-8 µm diam. located at the centre of the fungal plug below the stomatal cavity; the lower wall having a textura epidermoidea.

Conidiophores: absent. Conidiogenous cells: flaskshaped, sometimes with an extended neck which may become flexuous, thin-walled, hyaline, smooth 10-40 x 2-5 µm. Conidia: hyaline, aseptate, smooth, elliptical in side view, circular in end view, 2.5-4 x 1 µm. Trichogynes (Minter 1980): hyaline, aseptate, smooth, flask-shaped at the base, with a long, flexuous neck in their upper part, 20-40 x 3-5 µm, difficult to distinguish from the larger conidiogenous cells. Development of conidia: conidial ontogeny holoblastic by apical wall building; conidial maturation synchronous with conidial ontogeny; conidial delimitation by one presumably double septum; conidial secession schizolytic, presumably by fission of the double septum. Proliferation: sympodial, holoblastic by replacement wall building apices generally before secession of the previous conidium. Regeneration: not observed.

Teleomorph: ascocarps embedded in both sides of the leaf or in twigs, scattered, conspicuous, dull to shiny, dark grey to black whether wet or dry, with a rather diffuse outline, elliptical to elongated elliptical, the longitudinal axis usually being parallel with that of the leaf or twig, not markedly raising the surface of the substrate, 600-2000 x 400-600 µm, with a pale longitudinal central band when immature which bisects the black region of the ascocarp, when mature opening with a single longitudinal split positioned along the centre of the pale band; lips present, lining the split, colourless to fawn, well developed, persistent. In vertical section: the clypeus is subcuticular, although degraded tanned elements of the substrate epidermal or more often stomatal cells may sometimes be present separated from the cuticle and embedded in the clypeus. The clypeus is composed of cells 4-8 µm diam. with blackened walls forming a textura angularis c. 90 µm thick near the lips, becoming thinner towards the edge of the ascocarp where it extends as far as, and joins the basal wall. Some evidence of textura epidermoidea may be detected in the region of the clypeus adjacent to the cuticle; towards the region of the split, the walls of the clypeus cells become less blackened and merge gradually into the hyaline cells of the lips; the outermost of these lip cells are gelatinised and the lips are located in a broad area directly underneath the pale longitudinal band visible from the outside and are more extensive than in most other species of this genus; the basal wall in vertical section has a more or less smooth outline, being composed of cells with blackened walls forming a layer 1-2 cells thick adjacent to and above the degraded leaf mesophyll; the subhymenium is composed of hyaline cells, 4-6 µm diam., forming a textura angularis several cells deep from which the hymenium arises; a few larger but otherwise similar cells 8-12 µm diam., forming a similar textura angularis are sometimes seen inside the angle made by the junction between the clypeus and the basal wall. In horizontal section the textura epidermoidea formed by the cells of the clypeus adjacent to the cuticle is very prominent.

Paraphyses: hyaline, filiform, smooth, septate, 1.5-2.5 µm wide, as long as, or slightly longer than the asci, unbranched or occasionally dichotomously branched with one side maintaining the principal axis, embedded in a gel 1-2 [tm thick, tips sometimes slightly swollen or hooked, bases of adjacent paraphyses often linked by hyphal bridges (Morgan-Jones & Hulton 1979). Asci: cylindrical, long-stalked, with a rounded apex, inoperculate, not amyloid, 120-170 x 8-10 µm. Ascospores: hyaline, aseptate, filiform, smooth, 70-110 x 1.5-2.5 µm, enveloped in a mucous sheath 0.5 µm thick, arranged fasciculately, sometimes coiled in a helix, 8 per ascus.

Fungi which inhabit the leaves of many groups of conifers have now been intensively studied. However, little attention has as yet been given to the leaf mycoflora of the Araucariaceae, the family to which Agathis Salisbury belongs. A summary of the published work relating to this family (Butin 1981) revealed that the few studies which have been made principally concern the genus Araucaria Jussieu in South America. No species of Lophodermium has hitherto been recorded on any species of the Araucariaceae (Hunt & Ziller 1978), although their occurrence is common and well-documented on species in other families of conifers. Since individual species of Lophodermium on conifers are normally restricted in host range to a single genus, the absence of records of Lophodermium from the Araucariaceae is of itself strong evidence that the present fungus is a new species. This supposition is confirmed by examination of descriptions of Lophodermium species occurring on other conifers (Darker 1932, 1967; Tehon 1935; Hunt & Ziller 1978; Minter 1981; Minter & Sharma 1982), which revealed no similar species.

The average age of the trees in the area from which the specimens were obtained is 200-300 years. L. agathidis was observed fruiting on recently fallen dead leaves and twigs. Ascocarps were observed to release spores by violent dispersal under moist conditions. Peak dispersal occurred in November and December and declined towards the end of March. Neither ascocarps nor conidiomata were seen on living leaves attached to the tree. Although this information is insufficient to make any detailed analysis it suggests that the ecology of L. agathidis is probably similar to that of other members of the genus (Millar & Richards 1975, Minter & Millar 1980). L. agathidis is probably, for much of its existence, a leaf endophyte inhabiting apparently healthy leaves, and fruiting only after death of the host from some other cause.

In taxonomic terms the ascocarps of L. agathidis present no unusual features except for the very conspicuous pale band bisecting the dark areas of the clypeus. This pale band is the result of the development in this species of lips which are larger than in any other member of this genus known to us. In the conidiomata development of the conidiogenous cells and conidia is normal for anamorphs of Lophodermium. The most interesting feature is the use of the host stomatal aperture as an ostiole in a manner similar but not identical to that of the anamorph of Lophodermium baculiferum Mayr, where several ostioles are produced under the leaf cells surrounding each stoma (Minter 1980). The anamorph of L. agathidis is probably at present best placed in the form genus Leptostroma Fr., following Sutton (1980) and Minter (1980), but as noted by Minter & Sharma (1982) the generic concepts in this group of coelomycetes are most unsatisfactory and in need of thorough revision.