The autumn crocus is principally known as a very pretty flower that looks a lot like a true crocus, but blooms in autumn rather than spring and contains a deadly poison called colchicine, similar to arsenic in its effect. Since autumn crocus can be mistaken for wild garlic (ramsons), we are wisely warned away from it. Ancient and medieval herbologists were well aware of its toxicity, and many physicians were reluctant to use it for that very reason. However, some eminent physicians, including Dioscorides and Avicenna, prescribed it as a "chemotherapy" treatment for cancers. The results were not by any means miraculous, but these early experimenters had the right idea. According to John Riddle in "Ancient and Medieval Chemotherapy for Cancer," Isis 73 (3): 313-330:
Autumn crocus contains a sufficient concentration of colchicines for pharmaceutical efficacy, but on the basis of modern studies, we conclude that the drug would arrest tumor mitoses in man but would not produce a complete regression in a malignant tumor. An ancient or medieval physician, treating a malignant growth topically and possibly internally, would probably observe a beneficial response but not a “cure.”
Riddle's article is fascinating and goes far beyond the single example of autumn crocus to illustrate the remarkable sophistication of these often undervalued early researchers and practitioners:
A search through the leading pharmaceutical and medical authorities of the Greco-Roman, classical Islamic, and medieval periods reveals that they recommended many of the same natural sources as those for the compounds discovered in the 1960s and 1970s and currently utilized in cancer treatments.
But here is just a little more on the history of autumn crocus "chemotherapy" to show the level of scientific awareness with which ancient and medieval physicians approached its use:
In the first century Dioscorides (fl. ca. A.D. 50-79) employed a drug made from autumn crocus (Colchicum autumnale L.), the very plant investigated by A/ P. Dustin in 1938 as an antitumor agent. Dioscorides recommended that the plant (kolchikon) be “soaked in wine and administered to dissolve tumors (oidemata) and growths (phumata) not yet making pus.”
. . .
Mattaeus Platearius (d. 1161?) and Avicenna employed the plant in ways that suggest antitumoral activity. . . . Abu Mansur (fl. 968-977) said that the drug concocted from it is poisonous but dries up old sores. In light of this evidence one can conclude that prior to the thirteenth century autumn crocus was employed as an anticancer agent, but that its use was not widespread. The reluctance may have been due to the belief expressed by Hildegard [von Bingen], who said that it was more of a poison than a medicine.
Which of course it was, but all medicine is a balancing act between killing and curing, and the ancient and medieval doctors who used autumn crocus apparently knew what they were doing and got results that were promising enough for them to recommend its use to others.