Fruit bodies up to 17 x 4.5 cm, branched, arbuscular. Stipe discrete, irregularly flattened, up to 45 x 7 mm, purple-black ("aniline-black") when fresh, overlaid by a white "bloom" which persists on drying, pilose to matted-tomentose below, arising from a significant patch of mycelium on soil. Major branches 2-4, dull purple ("anthracene-purple"), smooth, irregularly flattened, drying to deep purplish grey; hymenium dull avellaneous ("light brownish drab"). Branches in 2-4 ranks, with sterile patches, ascending, flattened, vinaaceous purple ("dull Indian-purple") when fresh, drying to avellaneous, minutely papillose all over, with papillae more numerous and conspicuous upward; hymenium greyish violet ("vinaceous drab"), unilateral in many places; flesh off-white, appearing water-soaked, fibrous, solid outward, stuffed inward; axils narrowly rounded; internodes long, diminishing irregularly and little. Apices irregularly cristate to flagelliform, papillose, livid pink-flesh ("orange¬vinaceous", "Etruscan-red").

Tramal hyphae of branches up to 9 gm diam., hyaline, thin-walled, clamped, commonly guttulate, generally parallel, loosely packed. Hymenium thickening significantly; basidia 60-80 x 7-9 gm, narrowly clavate, clamped; contents multiguttulate, yellow-refringent; sterigmata 2, cornute, slender; effete basidia persistent; post-partal septation uncommon. Cystidia 80-160 x 9-10 Ftm, cylindrical, somewhat undulate, aseptate to occasionally 1-septate (septum clamped), refringent distally, in clusters of 10-50, emergent up to 100 gm, yellowish under bright field.

Spores 8.3-10.1 x 6.8-7.6 Ftm (E= 1.10-1.47; E^' =1.25; L'" = 9.04 gm), subglobose to broadly ellipsoid, thin-walled, hyaline; contents uniguttulate, the guttule almost filling the spore lumen; hilar appendix papillate, small.

Unfortunately, I have seen only one fruit body of this taxon, but its size, colour, and micromorphology are so unique that I propose a new species. Purple colours are not abnormal in Clavulina, for almost all those taxa whose fruit bodies are grey to mouse-grey have a violet to purple component to the colour. Clavulina purpurea is the only taxon I know in which the purple pigments are unmasked. It would appear that immature branches must be purple to reddish violet (upward) but must be masked by the grey hymenium which thickens gradually but significantly.

From characters used in the key above, Clavuhna purpurea would seem close to C. brunneo-cinerea, but C. purpurea shows a nearly black stipe and papillate upper branches. Fruit bodies of C. brunneo-cinerea are considerably bushier, with thicker, shorter stipe and more congested branches. Cystidia in Clavulina purpurea occur almost exclusively in large clusters so emergent as to be macroscopic. The clustering phenomenon is not unique, being shared by several other cystidiate taxa, but only C. hispidulosa has been reported as showing macroscopic clusters, although not of the magnitude of those found in C. purpurea.