Basidiome membranaceous, hymenium not easily separable. Subicular hyphae hyaline, thin- to thickwalled (up to 1 µm), 2.5-5.5 µm wide, without clamps. Cystidia hyaline, thin- to thick-walled (up to 1 µm), but thin-walled at the apex, cylindrical, flexuous, rarely with short branches (e.g., at the apex), 40-100 x 6-10.5 µm, with simple septa, encrusted with crystals over most of the length, but encrustation thin or absent at the apex. Basidia clavate 19-40 x 4-7 µm, occasionally repetitive, without clamps at the base. Spores hyaline, smooth, thin-walled, ellipsoid to cylindrical, 6.5-10 x 3-4.5 µm, not amyloid.

The species has been placed in Candelabrochaete because of its septate cystidia, the repetitive basidia and the simple-septate hyphae. It differs from the type species mainly by the denser texture (only young basidiomes have arachnoid parts, cf. PDD 12609), but fits in the genus in the broader sense of Burdsall (1984) and Hjortstam (1983).

Holotype: on Pittosporum eugenioides A. Cunn., New Zealand, Taupo, Mt Pihanga, coll. J.M. Dingley, Oct. 1949 (PDD 7363).

Basidiome pellicular to membranaceous, separable. Hymenial surface smooth to warted, cream-coloured to ochraceous, locally with orange spots. Margin fibrillose. Subicular hyphae thin- to slightly thickwalled, hyaline, 1.5-4 µm wide, with clamps at all primary septa. Hyphae often with crystals, sometimes completely ensheathed. lamprocystidia hyaline, somewhat thick-walled, heavily encrusted, 25-50 x 7.5-12 µm. Encrustation dissolving in KOH, revealing a rather thin-walled upper part of the cystidium, cylindrical, 3.5-4.5 µm wide. Leptocystidih hyaline, thin-walled, acute; 30-55 x 2.5-3.5 µm, sometimes with narrow apical protrusion. Basidia clavate to subcylindrical, 15-24.x 3-4 µm, with four slender sterigmata. Spores hyaline, smooth, thin-walled, obovoid, 4-55 x 2.5-3.5 µm, not amyloid.

The species is best classified in Ceraceomyces Jülich despite its cystidia and relatively thick hymenial layer. Encrusted cystidia are also present in C. cystidiatus (J. Eriksson & Hjortstam) Hjortstam, but here the hymenial layer is much thinner. In fact, C. cerebrosus may represent an intermediate between Ceraceomyces and Phanerochaete Karsten, sharing with the latter similar cystidia and a relatively thick hymenial layer. Traditionally, Phanerochaete is restricted to species without clamps or with occasional clamps (sometimes in whorls) on the subicular hyphae. It is remarkable that New Zealand has several species which are best classified in Ceraceomyces, but show strong tendencies towards other related genera. C. corymbatus (G.H. Cunn.) Stalpers and C. varicolor (G.H. Cunn.) Stalpers, for example, are intermediates between this genus and Ceraceomerulius Bond. and Meruliopsis Bond., respectively (if these genera are considered as distinct).

Holotype: on Nothofagus fusca (Hook. f.) Oerst., New Zealand, Westland, Staircase Creek, coll. S.D. Baker, Nov. 1954 (PDD 11871).

Basidiome annual, resupinate, effused. In section the structure is distinctly duplex, showing a loose white basal layer, a dense brown subhymenial layer and a cream-coloured hymenium. Hymenophore even, rarely warted, cracked, cream-coloured to ochraceous. Hyphae hyaline, thin- to somewhat thick-walled in the basal layer, 2-4(-4.5) µm, with distinct clamps. Vesicles stalked, in the basal layer up to 12 µm diam., in the upper layer up to 7(-8) µm diam., often with resinous contents. Clusters of crystals present in the trama, sometimes mimicking cystidia. Hymenial cystidia clavate, thin-walled, smooth or rarely finely encrusted, 30-50 x 5.5-6.5 µm, with hyaline or sometimes refractive material. Basidia narrowly clavate to nearly cylindrical, sometimes sinuous, 18-25 x 4-5 µm. Spores hyaline, thin-walled, smooth, ellipsoid to slightly flattened, 4.5-6(-6.5) x 2.5-3 x 3-4 µm, not amyloid.

The species belongs to Chondrostereum Pouzar because of the duplex structure, the vesicles, the hymenial structure and the narrow basidia. It is close to C. coprosmae (G.H. Cunn.) Stalpers, another species of this genus endemic to New Zealand.

Holotype: on Pittosporum eugenioides, New Zealand, Auckland, Kaiwaka, coll. J.M. Dingley, May 1949 (PDD 7351).

Basidiome perennial, membranaceous to subceraceous, cracked when dry. Hyphae hyaline, thin-walled, 1.5-3.5 µm, with clamps. lamprocystidia encrusted, conical to ovoid or ellipsoid, 20-35 x 10-18 µm. Crystals partly dissolving in 6% KOH, underlying structure thin- to somewhat thick-walled, hyaline, ellipsoid to ovoid, up to 14 µm wide. Gloeocystidia sulphobenzaldehyde positive, thin- to somewhat thick-walled, ellipsoid to narrowly ovoid, 18-40 x 10-15 µm, filled with yellowish refractive material. Some "gloeoplerous" hyphae also present, 4-5 µm wide. Occasionally a papillate structure was present in the hymenium,1.5-2 µm wide; this structure could not be found intact, so it is not known whether it is a second kind of gloeocystidium or a hyphidium. Basidia urniform, 15-18 x 4.5-5.5 µm, in the lower part up to 7.5 µm wide. Spores hyaline, subglobose, 5.5-7.5 x 4.5-6.5 µm, warty to echinulate, amyloid; warts and amyloidity disappearing with 6% KOH.

Gloeocystidiellum convolvens (Karsten) Donk is similar to this species in possessing both gloeocystidia and lamprocystidia. However, G. convolvens differs in the absence of clamps and the shape of the gloeocystidia.

Holotype: on Schefflera digitata J.R. & G. Forst., New Zealand, Coromandel, Whitianga Road, coll. J.M. Dingley, Aug. 1954 (PDD 13825).

Hyphae hyaline, thin-walled, 2-3.5 µm wide, septa without clamps. Gloeocystidia present but sometimes rare, acute, 20-40 x 4-6 µm, some weakly sulphobenzaldehyde-positive. Cystidia hyaline, thin- to slightly thick-walled, encrusted and then 30-45(-50) x 7-10(-15) µm, dextrinoid; cystidia fusiform and up to 6.5 µm wide without crystals. Basidia hyaline, thin-walled, clavate to urniform, 25-35 x 6.5-8 µm, (basal swelling up to 9 µm wide), with four sterigmata up to 8 µm long. Spores hyaline, thin-walled, subglobose to ellipsoid, often collapsed 7.5-9.5 x (5.3-)6-7.5 µm, echinulate, amyloid, appearing more or less smooth in KOH.

This description differs from that by Cunningham (1955, 1963) in reporting the presence of gloeocystidia and describing the basidiospores as amyloid and echinulate. The weakly sulphobenzaldehyde-positive reaction of gloeocystidia in the type specimen need not be characteristic of the species. The reaction to sulphobenzaldehyde is often indistinct in herbarium material as it depends on both age and method of preservation. It should be tested on fresh material.
The species is closely related to Gloeocystidiellum convolvens (Karsten) Donk, which also has encrusted cystidia, gloeocystidia, simple-septate hyphae and echinulate, amyloid spores. Gloeocystidiellum sacratum differs in the smooth hymenophore (only an occasional wart present), the scarce and small gloeocystidia, and larger basidia and basidiospores. For these reasons a new combination in Gloeocystidiellum Donk is proposed. The alternative would be a new genus for both species.
Burdsall (1985) placed the species in Amylostereum Boidin, because of "the hyaline, simple-septate hyphae, subulate reddish-brown pseudocystidia and amyloid spores with a granulose surface". This is rather surprising as all species of Amylostereum so far described have smooth, cylindrical basidiospores and clamps at all septa; most of them also possess thick-walled, brown basal hyphae. Gloeocystidiellum sacratum has hyaline cystidia, which become reddish brown in Melzer's reagent. Moreover the presence of gloeocystidia, noted by Taylor (1969), is not a typical character of Amylostereum. Finally, cultures of G. sacratum (CBS 796.86) display occasional septa with 1-3 clamps, a character known in Gloeocystidiellum (Boidin 1966), but not in Amylostereum.

Holotype: on Leptospermum scoparium J.R. & G. Forst., New Zealand, Waikato, Kawhia Harbour, coast, coll. J.D. Atkinson, Jan. 1953 (PDD 11845).

Subicular hyphae hyaline, thick-walled (up to 2 µm), cyanophilous, up to 11.5 µm wide. Cystidia hyaline, cylindrical, thick-walled, but wall becoming thinner towards the apex, septate with small prominent clamps, partly covered with a sheath of yellowishbrown crystalline material of about 1 µm thick, up to 500 x 7-10 µm. Basidia and other hymenial characters not seen. Spores hyaline, somewhat thickwalled, subglobose to ellipsoid, 4.5-6.5 x 4-5 µm (measured in Melzer's reagent and cotton blue in lactic acid), cyanophilous, not amyloid. In alkaline solutions the spores swell considerably, 4.5-7.5 (-8.5) x 3.5-4.5 X 4-6 µm.

Cunningham (1953, 1963) used the words "mucilaginous warts" in his description of the encrustation of the cystidia. This is understandable as the rather fine crystals are embedded in amorphous material; under pressure the encrustation breaks away from the cystidia as small plates. The encrustation dissolves in KOH.
Cunningham's measurements of the basidia (8-13 x 4-12 µm) seem to be an error. Although no basidia were recovered in our slides, it is clear from Cunningham's drawing and the description (subclavate, rarely cylindrical), that the actual width is much less than 12 µm.
The species strongly resembles Hypochnicium polonense (Bres.) Strid (=Kneiia polonensis Bres.), but differs in some minor characters. The encrustation of the cystidia of G. zealandica is darker than the nearly hyaline to pale yellowish crystals of H. polonense, and the spores are slightly smaller and more broadly ellipsoid. However, Eriksson & Ryvarden (1976) remarked in their description of H. polonense "a remarkable variation in the size of basidia and spores, and also a variation in spore shape, which may vary from almost allantoid to broadly ellipsoid in a single collection. However, these variations do not seem to be of any taxonomic significance." These observations are confirmed, although some collections had a stronger tendency towards cylindrical spores than others. We never observed allantoid spores; similar spores always turned out to be collapsed. An interesting observation is that in the type specimen of G. zealandica, and in most specimens of K. polonensis, spores of a Tomentella or Thelephora were seen.
Kneiffia polonensis (Fig. 2) has been placed in Hyphoderma Wallr. emend. Donk and more recently in Hypochnicium J. Eriksson but does not belong in either genus because of the texture of the basidiome, the wide subicular hyphae and the septate subicular cystidia.
Species of Candelabrochaete Boidin have similar cystidia, but they are formed by prolongation of the main axis of a basidial cluster. Moreover, Candelabrochaete species have hyphae which are not cyanophilous and are typically devoid of clamps.
We accept Jülich's combination Gyrophanopsis zealandica. As Kneiffia polonensis is only slightly different, it is included here also: Gyrophanopsis polonensis (Bres.) comb. nov. (basionym: Kneiffa polonensis Bres. - Ann. Mycol. 1: 102. 1903). Hyphodermopsis Jülich, typified by Kneiffia polonensis, is thus considered a synonym of Gyrophanopsis Jolich.

Holotype: on Prumnopitys taxifolia (D. Don) Laubenf. (=Podocarpus spicatus R. Br.), New Zealand, Auckland, Te Whaiti, coll. J.M. Dingley, Jun. 1951 (PDD 11449)

Hyphae hyaline, thin-walled, 2-4(-5) µm wide, with clamps at all septa. Vesicles abundant, 10-18 µm diam., stalked, not cyanophilous. Glococystidia not seen in the type specimen (which is in good condition), but present in some other collections (e.g., PDD 13684). Cystidia hyaline, conical, thick-walled, 60-120 x 12-17 µm, acute, but under the encrustation obtuse. Basidia urniform, 30-45 x 6-8 µm. Spores hyaline, thin-walled, smooth, cylindrical to curved or sigmoid, 9-14(-16) x 3-4.5 µm, not amyloid.

The species clearly belongs to Hyphoderma Wallr. and is close to H. puberum (Fr.) Wallr., differing in the presence of vesicles and in the shape and size of the spores. The inconstant occurrence of gloeocystidia is also known for H. puberum (Eriksson & Ryvarden 1975).

Holotype: on Leptospermum scoparium, New Zealand, Auckland, Waitakere Ranges, Anawhata Road, coll. J.M. Dingley, Sep. 1949 (PDD 7348).

Basidiome membranaceous, white to creamcoloured. Hyphae thin-walled, clamped at all septa. Cystidia strongly projecting, subulate, thick-walled, multiradicate (often biradicate) at the base; wall covered with crystals, inner layer(s) dextrinoid at the base; 100-160 x 11-21 µm, sometimes with one or more hyphae which may ensheath the cystidium. Cystidioles originating from subicular hyphae, thinwalled, subulate,11-20 x 2.5-3.5 µm, the upper half or only the apex covered with coarse, sharp crystals. Basidia barrel-shaped to ovoid, often constricted towards the apex, with (2-)4 sterigmata,15-25 x 810 µm, at least some basidia arising laterally. Spores hyaline, vemuculose, globose to subglobose, 5.5-11.5 x 5-10 µm, not amyloid. Ornamentation consisting of circular, flattened warts, visible (although often faintly) in Melzer's reagent but dissolving in KOH.

Cunningham (1955,1963) reported that spores of P. gladiola were smooth, possibly because he examined material only in KOH. Under SEM, some variation in spore ornamentation is evident between different collections, e.g., spores from the holotype (Fig. 8) have smaller and more closely packed warts than those from a paratype collection (PDD 11444, Fig. 9).
The species belongs to Litschauerella and is very close to L. clematitis (Bourdot & Galzin) J. Eriksson & Ryv., the only species with globose to subglobose spores. Litschauerella gladiola has been considered a synonym of this species by Weresub (1961) and Oberwinkler (1965, sub Peniophora abietis (Bourdot & Galzin) Botudot & Galzin). It differs, however, from this species by the presence of encrusted cystidioles, larger cystidia and the ornamentation of the spores.
Spore ornamentation in the complex around L. clematitis is rather variable, as demonstrated by Eriksson & Ryvarden (1976). They described the ornamentation of L. clematitis as widely spaced 4sided, flat-topped warts, but also described a specimen of an, as yet, undescribed species of Litschauerella, having spores with irregularly arranged, thin, erect, rectangular plates, and another) with nearly smooth spore walls.
Spore size shows wide variation within L. gladiola. Spores from the holotype (6.5-11.5 x 6.5-10 µm, Fig. 8) were on average larger than those in other PDD collections (e.g., PDD 11444, 5.5-9 x 5-8.5, Fig. 9; spores also showed larger warts than those in the holotype). However, spores in all collections seen were globose to subglobose, and collections could not be distinguished by other morphological features.

Holotype: on Brachyglottis repanda Forst. & Forst. f., New Zealand, Auckland, Taneatua Reserve, coll. G.H. Cunningham, 23 May 1952 (PDD 11484).

Basidiome membranaceous, white. Hyphae thinwalled; clamped at all septa, encrusted in the subiculum. Cystidia projecting, subulate, thickwalled (wall crystal-covered), swollen at the base, c. 130-185 x 10-15 µm, with climbing hyphae. Mature basidia not seen. Spores hyaline, verrucose, ellipsoid to obovoid, 6-7.8 x 5-5.8 µm, not amyloid. The ornamentation of the wall consists of irregularly arranged, thin, erect, angular (often rectangular) projections; it is distinct in Melzer's reagent but only faintly visible in KOH suggesting partial dissolution.

The species is close to Litschauerella gladiola, as recognised by Cunningham (1955, 1963) and Weresub (1961). The rather small ellipsoid spores and the relatively coarse ornamentation distinguish it from both L. clematitis and L. gladiola. The latter species also has encrusted cystidioles. Weresub (1961) suggested that the poor condition of the holotype of L. hastata "could be ascribed to growth of P. clematitis under unfavourable circumstances of some kind". However, a more recent collection of L. hastata (PDD 48739, Fig. 11), agrees with the holotype (Fig. 10) in spore characters. Studied with SEM, the erect, angular projections of L. hastata (Fig. 10-11) are different from the rounded warts on spores of L. gladiola and also from the cubical projections of L. clematids spores. It may be identical with Litschauerella sp. Bourdot 32350 (= L. Corbière 5) as described by Eriksson & Ryvarden (1976).

Holotype: on Griselinia lucida Forst. f., New Zealand, Auckland, Piha, Glen Esk Valley, coll. J.M. Dingley, 12 May 1951 (PDD 11442).

Basidiome resupinate, effused, annual, arachnoid to submembranaceous. Hymenium even. Hyphae hyaline, thin- to rarely somewhat thick-walled, locally inflated, with clamps. Leptocystidia straight, cylindrical, protruding, originating from basal hyphae. lamprocystidia small, originating in the trama, encrusted with coarse crystals. Basidia hyaline, thin-walled, subclavate, terminal, with four sterigmata. Spores hyaline, thin-walled, not amyloid. Type species: Peniophora verecunda G.H. Cunn.

Basidioma resupinatum, effusum, annuum, arachnoideum vel submembranaceum. Hymenium leve. Hyphae hyalinae, tenui- vel raro crassi-tunicatae, fibulatae, interdum inflatae. Leptocystidia cylindrica, exserta, oriunda ex hyphis basilaribus. lamprocystidia parva, crystallis crassis incrustata, ex subiculo oriunda. Basidia hyalma, tenui tumcata, subclavata, temunaha, 4 sterigmata gerentia. Sporae hyahnae, tenui tunicatae, inamyloideae.

Etymology: palus (pole) and -fer (bearer)

Species typica: Peniophora verecunda G.H. Cunn.

Basidiome effused, arachnoid to submembranaceous. Subiculum loose, scanty. Basal hyphae thin- to rarely somewhat thick-walled, up to 4.5(-7) µm wide, sometimes irregularly inflated, with small, abrupt clamps. Vesicles 3.5-4.5 µm wide. Encrusted hyphae present in subiculum; the encrustation is irregular, either apical or surrounding intercalary parts of hyphae. Thin-walled cystidia with large, rodshaped crystals present in the subiculum, originating from parallel basal hyphae or directly from the wood, 15-32 x 3-5 µm. Leptocystidia cylindrical, smooth, arising from the subhymeniunn or subiculum, 40-100 x 4.5-6 µm. Basidia rarely recognisable, subclavate, c. 15-20 x 4 µm, with four sterigmata. Spores hyaline, thin-walled, smooth, broadly ellipsoid, 4-6 x 2.5-4 x 3.5-4 µm, not amyloid.

This species does not fit in any described genus. The structure of the basidiome suggests Botryobasidium Donk, but the morphology of the hyphae, vesicles and encrusted cystidia do not correspond to this genus. Grandinia Fr. has different hyphae and cystidia. In Fibulomyces Julich or Leptosporomyces Julich vesicles and encrusted cystidia are unknown. Consequently a new genus [Palifer] is described:

Holotype: on Dacrydium cupressinum Lamb., New Zealand, Taupo, Hauhungaroa Range, coll. J.M. Dingley, Mar. 1953 (PDD 12513).

Subicular hyphae hyaline to yellowish, thin- to thick walled (up to 2.5 µm), finely asperulate, 10-18 µm wide. Subhymenial hyphae hyaline, thin-walled, 6-12 µm wide, cyanophilous. Basidia collapsed. Spores hyaline, smooth, ovoid from above, navicular in side view, 5.5-7.5 x 2.8-3.5 x 3-4 µm, somewhat thick-walled with prominent apiculus.

The species is indistinguishable from Botryobasidum pruinatum. The spore size is much smaller than Cunningham's (1953) measurements (8-11 x 3.5-4.5 µm). The basal hyphae are somewhat less thick-walled, paler and more finely asperulate than typical European'specimens, but these differences are herein considered to be within the natural variation of the species.

Holotype: on Pseudopanax arboreus (Murr.) Philipson, New Zealand, Taranaki, Mt Taranaki coll. GH. Cunningham, Feb 1952 (PDD 11296).

Lamprocystidia more or less conical, thick-walled, encrusted, 35-115 x 12-20 µm, hyaline, rarely basally yellowish. Encrustations covering 30-80% of the cystiddal surface. Spores hyaline, thin-walled; cylindrical, 9-11.5 x 4-5 µm.

Peniophora coprosmae is a good species of Peniophora s. str. and closely related to P. incarnate (Pers.:Fr.) Karsten. It differs from this species by a more strongly developed and denser subiculum, and the concentration of both lamprocystidia and gloeocystidia in the hymenium, as already indicated by Cunningham (1963).

Holotype: on Coprosma robusta Raoul, New Zealand, Wellington, Lake Papaitonga, coll. G.H. Cunningham, Oct. 1930 (PDD 3719).

Spores hyaline, thin-walled, smooth, ellipsoid, 5.5-8 x 3.5-4.5(-5) µm, not amyloid.

The species was well described by Cunningham (1963) and belongs to Peniophora sensu stricto. Its closest relative is probably P. versicolor (Bres.) Sacc. & Sydow, which differs mainly by its larger spores.

Holotype: on Laurelia novae-zelandiae A. Cunn., New Zealand, Bay of Plenty, Mt Te Aroha, coll. G.H. Cunningham, Sep.1954 (PDD 13828).

Hyphae hyaline, thin- to slightly thick-walled, 2-4 µm wide, clamped at all septa. Cystidia thickwalled, encrusted, 20-45 x 5-8.5 µm, when naked up to 5 µm wide. Basidia cylindrical, 20-35 x 5-6.5 µm. Spores ellipsoid to cylindrical, smoothwalled, not amyloid, 7-9 x 3.5-4 µm.

The type material is in poor condition, with hyphal characters obscure and spores absent. Other PDD collections also were found to be sterile. Two fresh collections on Podocarpus hallii made at the type locality (PDD 48759, 48760) provided the basis for the above description of hyphal septation and spores. The species was well described by Cunningham (1955, 1963), although cystidia did not appear to be arranged in rows. The species belongs to Phlebia Fr. because of its waxy consistency when fresh, clamped hyphae, ellipsoid spores, and narrowly clavate basidia. PDD collections were annotated by H.H. Burdsall as Phlebia totara but this combination was not published (Burdsall, pers. comm.).

Holotype: on Podocarpus hallii Kirk, New Zealand, Taupo, Mt Ruapehu, Silica Springs Track, coll. S. D. Baker, Jan. 1954 (PDD 13573).

The spores of S. otagense are slightly smaller than Cunningham's measurements (5-7 x 3.54.5 µm). This brings it in the range of S. diademiferum (Bourdot & Galzin) Donk, a widely distributed species growing on a large variety of woody substrates. Eriksson et al. (1984) described the spores of S. diademiferum as ovoid to subglobose, (3.5-)4-5.5(-6) x (2.5-)3-3.5 µm; the spores from a French collection of S. diademiferum (LY 3.454) measured 4-5 x 2.5-3.5 µm. Sistotrema otagense differs from S. diademiferum only in the broader basal hyphae and the slightly different shape of the spores.

Holotype: on unknown host, New Zealand, Otago, Dunedin, Morrison's Creek, coll. G.T.S. Baylis, Jun. 1952 (PDD 11581).

Basidiome effused, thin-arachnoid, white. Basal hyphae hyaline, thin-walled, 2-3 µm wide, with clamps. Cystidia thin-walled, acuminate, up to 110 µm long, apically 2 µm wide, basally up to 7.5 µm wide, covered with plate-like crystals. Basidia thin-walled, in dense clusters. Spores hyaline, thin-walled, smooth, ellipsoid to allantoid, 7-9 x 3.2-4.2 x 3.8-5 µm, not amyloid.

The species belongs to Subulicystidium Parm. The nature of the cystidia is similar to that of S. longisporum (Pat.) Parm. as described by Jülich (1975).

Holotype: on Rhopalostylis sapida Wendl. & Drude, New Zealand, Auckland, Waitakere Ranges, Cascades, coll. S.D. Baker, Apr. 1954 (PDD 13816).

Lyocystidia radicate, tapering, with an inflated apex, 35-60 x 3.5-4.5 µm, apex 4.5-6 µm diam. Mature basidia not seen. Spores globose to subglobose, 3.2-5 x 3.5-4.7 µm, not amyloid.

We concur with the descriptions of this species given by Cunningham (1955, 1963) and Weresub (1961). The delicate hymenophore and substrate of the holotype make observation difficult; PDD 11483 is the only New Zealand collection of T. thermometrus. The species differs from T. accedens (Bourdot & Galzin) Donk by the (sub)globose spores and smaller cystidia.

Holotype: on Metrosideros robusta A. Cunn., New Zealand, Bay of Plenty, Hick's Bay, coll. G.H. Cunningham, May 1952 (PDD 11483).

Hyphae hyaline, much branched, short-celled, clamped at all septa. Lyocystidia strongly projecting, radicate at the base, gradually, tapering, broadly rounded at the apex and covered by a large umbrella-shaped crystalline head, 6-8.6 µm across, not soluble in KOH; cystidium wall hyaline, smooth, not amyloid, unevenly thick-walled up to 3.7 µm wide; wall frequently asymmetrically thickened, dissolving in 10% KOH; cystidia 38-70 x 6.8-10 µm. Miniature cystidia of the same shape, possibly juvenile forms, up to 4-7 x 6 µm, with umbrella-shaped crystalline head 5 µm across. Basidia barrel-shaped, with four sterigmata, 12.5-17.5 x 5-7 µm. Spores hyaline, smooth, oblong-ellipsoid, 4.5-6.5 x 2.5-3.5(-4.5) µm, not amyloid.

The conspicuous capitate lyocystidia of T. umbraculus are found in only two other species, T. hamatus (H.S. Jackson) Donk and T. corneri Jülich. Both T. umbraculus and T. hamatus have an effused, even hymenophore, whereas that of T. corneri is odontioid to hydnoid (Jülich 1979). Oberwinkler (1966) and Jülich (1979) considered T. umbraculus and T. hamatus conspecific, but Weresub (1961) found that the holotypes of these two species could be distinguished by spore shape; spores of T. hamatus were described as "broadly ellipsoid, 5.5-7(-8.2) x 4-5.5 µm". Weresub (1961) and Cunningham (1955) also noted that basidia of T. hamatus are larger (up to 30 x 8.5 µm) than those of T. umbraculus. Furthermore, the cystidia, while similar with the conspicuous crystalline head, apparently differ in wall thickening; in T. hamatus the walls were rarely up to 2-3 µm thick and not reported to be asymmetrically thickened (Weresub 1961).
In confirming the careful observations made by Weresub (1961) for T. umbraculus and accepting her description of T. hamatus, we consider that separation of the two species is justified on the basis of morphological differences in cystidia, basidia, and spores.

Holotype: on Pseudowintera colorata (Raoul) Dandy, New Zealand, Stewart Island, Horse Shoe Bay, coll. J.M. Dingley, 17 Feb. 1954 (PDD 13660).