Gregarious to fasciculate in mainly rich, calcareous Fagus forests. Also reported with other trees, both in deciduous and coniferous forests. According to the Norwegian Mycological Database recorded once in Norway (Troms). To my knowledge there is no voucher material, and I think the record is doubtful. It is more likely to be found in beech forests in Southern Norway, where it was found for the first time in 2010.
Pileus 15-45 mm across, campanulate, becoming shallowly conical, generally with a fairly small umbo and a raised concentric zone near the margin or halfway the apex, translucent-striate, hygrophanous, smooth, glabrous, somewhat greasy-shiny when moist, at first dark brownish violet to reddish violet (except at the umbo which is watery grey-violet or even ochraceous-tinted), fading to pale straw yellow-violet, pale greyish violet or pale greyish pink, with or without brownish shades,the umbo often becoming dingy whitish and contrasting with the concentric zone and the margin which retain their original colour longer. Odour sweetish fragrant mixed with a component resembling the scent of a wooden cigar box (the mixture rather reminiscent of incense), but raphanoid when cut or bruised. Taste raphanoid. Lamellae 24-32(-36) reaching the stipe, narrowly adnate, smooth to veined, becoming dorsally intervenose, like the pileus dark brownish violet to reddish violet, turning more purplish brown to reddish brown with age, the edge convex, slightly eroded, whitish. Stipe 40-80(-110) x 1.5-4.5(-6) mm, hollow, firm but brittle, equal for the greater part, somewhat broadened below, straight, geniculate below, terete or somewhat compressed (and splitting lengthwise), smooth, pruinose to minutely puberulous at the very apex, glabrous and shiny farther down, concolorous with the young pileus, with age becoming increasingly more brown (pinkish brown to dark brown), the base covered with long coarse, flexuous, yellowish fibrils.
Basidia 24-30 x 6.5-7 μm, slender-clavate, 4-spored, with sterigmata 5.5-7 μm long. Spores 7.4-9.8 x 3.6-5.4 μm, pip-shaped to somewhat irregularly shaped, smooth, amyloid. Cheilocystidia 22.5-60(-80) x 3.5-20 μm, forming a sterile band, clavate, fusiform or subcylindrical to more irregularly shaped, smooth. Pleurocystidia scarce, similar. Lamellar trama dextrinoid. Hyphae of the pileipellis 1.5-2.5 μm wide, smooth, not embedded in gelatinous matter. Hyphae of the cortical layer of the stipe 1.5-3 μm wide, smooth, caulocystidia 3.5-10.5 μm wide, slender-clavate to cylindrical, smooth. Clamp connections present in all elements.
The macroscopic description has been taken from Krieglsteiner & Schwöbel's account. The microscopic details are based on Maas Geesteranus (1989 b).
Mycena diosma seems to have been accepted as a separate species by European mycologists (e. g. Maas Geesteranus 1989 b, Robich 2006, Emmet et al. 2008), although it is not much separating it from the highly variable M. pura. In trying to establish it as a species of its own, Krieglsteiner & Schwöbel (1982) pointed at its vivid colours, especially at the lamellae, and the peculiar smell. There is, however, no essential microscopical difference between M. diosma and M. pura.
In his key Maas Geesteranus (1989 b:488-489) separated the two taxa as follows:
- M. pura - lamellae white to somewhat coloured, but always pale.
- M. diosma - lamellae dark brownish violet to dark violet.
The other important feature is the smell, which has been descriped as sweetish, mixed with a scent of cigar box, but raphanoid when cut or bruised.
Phylogenetic analyses by Harder et al. (2010) clearly supported M. diosma as a distinct species. However, the morphological species recognition criteria described by Krieglsteiner & Schwöbel (1982) - dark purple lamellae and cigar box-like smell - are unreliable. Their results showed that neither violet lamellae nor the cigar box-like smell is always present. They instead suggested that the species can be morphologically recognized on a combination of amyloid spores and few, if any, pleurocystidia. The smell and the lamellae colour can be invoked as supportive characters.
Further images on the web: