This species is common throughout New Zealand and probably recognised by Colenso (1890) and Massee (1898) as Agaricus (Pholiota) pudicus. Indeed Colenso's record of Pholiota heteroclita (Fr.) Quel. may refer to the same species but the material (b 624) in K lacks basidiospores. It is an excellent edible mushroom very closely related to A. cylindrica ( DC.: Fr.) Maire, indeed at the start of the present study following Singer (1969) the present authors included it within their circumscription of A. cylindrica.

In addition Stevenson (1982 a & b) describes the same species on Alectryon excelsum Gaertn., Carpodetus serratus Forst., Corynocarpus laevigatus Forst., Dysoxylum spectabile Blume, Hoheria sp., Laurelia novaezelandiae A. Cunn., Nothofagus solandri Oerst. and Plagianthus betulinus A. Cunn. To this list can now be added Erythrina caffra. Leptospermum ericoides. Macropiper excelsum. Magnolia grandiflora Linn. (Univ. of Auckland), Metrosideros excelsa. Populus italica (Duroi) Muench. (Wynyard Car Park, Univ. of Auckland), Rhododendron and Senecio grandifolius Berg.

Thus A. parasitica has now been found on both native and introduced trees and would appear to be far commoner than A. cylindrica is in Europe which, although it has been recorded on several hosts, is much more restricted in its host preference. One record exists in New Zealand for A. parasitica on Ulmus glabra Huds., a host on which A. cylindrica is found in Europe. Illustrations of this species occur in Taylor (1981; 1983).

A. parasitica (as A. cylindrica) from Australia is illustrated by Macdonald and Westerman 1979 and Fuhrer 1985. Dr Fuhrer kindly sent a collection to Edinburgh for study and it confirmed our identification. Material examined: on Atherospermum moschatum. Talbot Drive, near Marysville, Vie., Australia, v 1985, B. Fuhrer B236. This may be the first record of A. parasitica in Australia, where it appears to be not uncommon in Victoria at least, and is probably widespread.

Singer (1969) synonymised Agaricus crassivelus Spegazzini (in Ann. Soc. Cienc. Argent. 9: 279, 1880), Pholiota formosa Spegazzini (in Bot. Acad. Cienc. Cordoba 28; 311, 1926) and Pholiota impudica Spegazzini (in Bol. Acad. Nac. Cienc. Cordoba ii» 414, 1889) with A. cylindrica (as A. aegerita) and recognised considerable variation within this taxon. He suggested that several taxa may ultimately be described in the complex. A. parasitica is just one such taxon; unfortunately during the present study Spegazzini's types were not examined and so it has not been possible to compare the microscopic characters demonstrated in our material with those published for the S. American taxa. However, collections from Brazil examined by one of us (RW) demonstrate that quite a different fungus is at least present there. Agaricus phylicigenus Berkeley from Tristan da Cunha is also claimed to be A. cylindrica (Singer, 1969), but the type in K is so badly preserved that it does not help to clarify the situation.

We have compared our material with the following collections of A. cylindrica from Europe (all in E) - The Netherlands; on Populus. Italy: on Ficus, Genova, Erbar Crittogam. Ital., Ser. II. British Isles: Petersham, Surrey, 27 v 1908, legit Hartley Smith; on Populus, Hereford, Ross-on-Wye. 29 x 1969, Orton 3503; on Acer campestre L., Archer Wood, Coppingford, Huntingdonshire, 5 v 1973, legit Sheila S Wells; on Populus, Offord, Huntingdonshire, 7 vii 1974, legit S. Wells; on Sambucus, Castor Henglands NNR, Cambridgeshire, 3 x 1976, legit S. Wells. Other material from the British Isles was examined during the preparation of the British Fungus Flora (Watling, 1982) and the data obtained there from has been used in this study.

In the Dutch and some English material on Populus the pleurocystidia are very rare, and the cheilocystidia vesiculose to ventricose and never with a long appendage. This is in fact parallel to material collected in Kashmir (on Populus alba L., Gulmarg, 28 ix 1978, Wat. 13067 in E & Abraham - 6RRL 7748 in E); see Watling & Gregory, 1980 and Watling & Abraham (ined.). In those British collections with abundant pleurocystidia these are always clavate, although often with a long pedicellate base, and subhymenial or even tramal in origin. In this way they resemble collections from Afghanistan (in garden, Kabul, 12 v 1951 legit Gehli; on Populus pyramidalis Rozier (= P. italica (Duroi) Muench) Kabul, 21 v 1951; both det. R. Singer in MICH), although the basidiospores of these agree more with A. parasitica (11-12 x 6-7 µm).

Thus whereas the European collections appear to be heterogeneous, those from New Zealand are uniform and characterised by the following cluster of characters;- relatively large basidiospores with small, although quite distinct germ-pore, elongate cheilocystidia and abundant development of pleurocystidia with long apical appendage.

Taylor 867 is notable in showing an excellent development of hyphal material from its bivelangiocarpic growth in parallel to the condition described for A. cylindrica by Reijnders (1979); however, this collection consists of immature basidiomes, judging from lack of spore-development; the basidiospores available are in fact comparatively small.

Some variation is found in the detailed structure of the pileipellis in the collections of A. parasitica but this simply reflects the state of maturity. At first the pileus is covered in hyphal strands under which a palisadoderm is formed; as the pileus expands these hyphae become intermixed with ellipsoid-pedicellate to spheropedunculate cells and at maturity are even joined by hyphae derived from the pileus-trama. This may explain the variation found in A. cylindrica where rugulose and smooth-capped forms occur but more critical notes are required on individual collections to ascertain the significance of such a character and whether it is correlated with other micro-data.

The variation in basidia, cystidial morphology and their abundance in Brazilian collections examined by one of us (RW) indicates that almost totally 2-spored forms are common there, 2- and 4-spored basidia were found on many of the New Zealand collections, e.g. Taylor 867, and in European collections from the Netherlands and from England (Orton 3503). Singer (1969) indicates that similar variation is found in the S. American collections he has examined, he also indicates that within these same collections of A. cylindrica (as A. aegerita) he found populations in which the basidiospores are "distinctly truncate whereas in others they are not, populations in which the pleurocystidia are acute and others in which the pleurocystidia are obtuse ventricose-ampullaceous, and populations which are exclusively 2-spored and others exclusively 4-spored.

As the basidiospores of members of the A. cylindrica group germinate so readily and the axenic cultures prepared fruit readily these agarics offer ideal subjects for the study of variation and interfertility. Esser and his colleagues (Esser, Semerdzieva & Stahl, 1974; Esser & Meinhardt, 1977; Meinhardt & Esser, 1981) have demonstrated the ways by which variability in European isolates can be studied, indeed they have analysed some 2-spored forms. Thus the conspecificity or autonomous nature of the various isolates in the A. cylindrica complex could be determined in parallel to the recent techniques used for analysing ranges of Armillaria isolates. A. cylindrica is in fact cultivated commercially in Europe; see Ferri, 1973.