Several fire ecologists at California State University at Los Angelos have been working extensively on identifying the chemicals in smoke that promote germination. In a research paper in the journal ecology (Ecology 79(7)1998 pp 2330-2336.), they found nitrogen dioxide as the main component of smoke which was involved in germination. Nitrates and nitrites had no effect. The chaparral species (annuals and shrubs) that they tested had complicated requirements. Some germinated with smoke alone, some required smoke plus stratification, and a few required abraison of the seed coat also. They were able to exactly duplicate results of smoke by substituting nitrogen dioxide in their trials. At the botanical garden here we use Kirstenbosch smoke papers for much of the South African seed we germinate and find them important for the germination of ericas, proteas and the grass-like members of the Restinonaceae. Restios for us will not germinate without smoke and protea germination is greatly inhanced. Bulbs on the other hand are a little more forgiving and we have had reasonable germination, especially of our own freshly collected bulb seed, without smoke . We still use smoke on stored bulb seed just in case it really does help. I've never had enough seed to do a true controlled test on bulb seed. Don Mahoney, San Francisco Botanical Garden