Comprehensive surveys designed to study the distribution of alkaloids in related plant families or within the genera and species of a given family are faced with the difficulty of collecting a sufficient number of representative genera or species for statistically significant conclusions. An obvious solution lies in the testing of herbarium material. Webb used the method several years ago and discussed the limitations of it in some detail (1, 2); he found that well preserved herbarium specimens containing relatively stable alkaloids gave positive tests after periods of storage of up to 125 years. In another experiment, one of us was able to detect trace amounts of alkaloids in plant material which had been stored under ideal conditions for approximately 1300 years (3), and other workers have reported success in the study of certain non-alkaloidal plant constituents in herbarium material (4,5). A recent editorial review of the subject (6) prompted us to record how we discovered and developed a "new" alkaloidal family of plants by this means.