This family of about 1100 species belonging to 85 genera is largely native to the tropics and sub-tropics, but some species occur naturally in temperate regions. Many of the plants are bulbous; others have rhizomes.
A large number of species in several genera are widely cultivated for their attractive flowers. These include species of Boophone Herbert, Crinum L., Haemanthus L., and Pancratium L. The genus Narcissus L., which includes the popular daffodils, narcissi, and jonquils, is very extensively cultivated. This results in frequent and prolonged contact by workers in the bulb and cut flower industries, and to a lesser extent by amateur gardeners and housewives. Somewhat less commonly encountered by gardeners, florists, and the like are various large flowered cultivars belonging to the genus Hippeastrum Herbert. The snowdrop (Galanthus nivalis L.) is also included in this family.
Narcissus species have been cultivated for over 300 years, as have a number of naturally occurring hybrids. Deliberate hybridisation was started in the 19th Century by English gardeners and has since been carried out on such a large scale that the parentage of many popular cultivars is uncertain (Gorer 1970).
The bulbs of Boophone, Crinum, Hippeastrum, Narcissus and of other genera are poisonous on ingestion (Watt & Breyer-Brandwijk 1962, Lewis & Elvin-Lewis 1977). This is probably the case for most species in this family, and is attributable to their content of alkaloids (Fuganti 1975).
Many reports describe skin irritation following contact with members of this family. Primary skin irritation may be caused in part by minute needle-like crystals of calcium oxalate, known as raphides, and in part by alkaloids present in the sap. Some of the alkaloids are also sensitisers.
This species occurs naturally in southern Africa but is widely cultivated. The genus is considered to be monotypic (Willis 1973).
The common name amaryllis is also applied to various large flowered cultivars belonging to the genus Hippeastrum, and to a lesser extent to Sprekelia formosissima Herbert, Vallota purpurea Herbert (syn. V. speciosa T. Durand & Schinz), and to certain species of Brunsvigia Heister, Crinum, Lycoris Herbert, Nerine Herbert, Sternbergia Waldst. & Kit., and Zephyranthes Herbert.
Gardner & Bennetts (1956) include "Amaryllis species" in a list of plants known or suspected of causing dermatitis.
This and other species of Boophone are natives of southern and eastern Africa, but may occasionally be found in cultivation. Boophone disticha is very poisonous if eaten, and is irritant to the skin (Kingsley 1967). Watt & Breyer-Brandwijk (1962) provide much support for the poisonous nature of the bulb in particular, but not for its supposed skin irritant properties. They do, however, record that the bulb has been used in southern Africa for the external treatment of various skin conditions.
The roasted bulb has been used in India as a rubefacient (Watt & Breyer-Brandwijk 1962, Quisumbing 1951).
The bulbs will reportedly blister the skin of cattle (Hurst 1942).
An aqueous extract of the leaves has produced a positive patch test reaction in a gardener (Bleumink & Nater 1974).
At least some species are irritant (Burkill 1935). The positive patch test reported by Agrup (1969) to the leaf of an unnamed cultivar is difficult to interpret.
von Reis Altschul (1973) found an herbarium note stating that the plant has been reported as very irritating to the skin.
The vernacular nomenclature of the various species is confusing. The term daffodil is commonly used for trumpet narcissi, especially Narcissus pseudonarcissus L.; the term narcissus is used for the remainder, but without consistency. The term jonquil is applied just as inconsistently to Narcissus jonquilla L., and to a number of wild daffodils, and to others thought to resemble them.
Walsh (1910) suggested that the irritant properties of the bulbs could be ascribed to the needle-like crystals of calcium oxalate found in many of the species. The leaves and stems are also mildly irritant.
Many authors refer loosely to "daffodils" and "narcissi" and thus little information is available concerning the sensitising capacity of the different species. However, N. jonquilla as well as N. poeticus and N. tazetta can sensitise (Stryker 1936, Agrup 1969). According to Klaschka et al. (1964), dermatitis from Narcissus is probably always caused partly by an allergic mechanism and partly by irritant effects, the clinical picture depending on the predominance of one or other mechanism and on the mode of exposure. Walsh (1910) noted that dogs also are sometimes affected when running amongst the flowers.
Nurserymen and pickers have repeatedly noted that some cultivars provoke dermatitis more readily than others. The cultivars Camparelle, Ornatus, Gloriosa, Scilly White, Grande Monarque (Walsh 1910), and Actaea and Princeps (Rook 1961) have proved particularly troublesome. Only a large and experienced grower can advise on the relative sensitising capacities of the cultivars in favour at a particular time and place.
The nature of the allergen is unknown (Hjorth & Wilkinson 1968). It is not present in an ether extract but is present in ethanol, acetone, and water extracts (Staines 1958, Bleumink & Nater 1974).
Narcissus absolute, a fragrance raw material prepared from the flowers of various Narcissus species (notably N. poeticus and N. tazetta), was found to be slightly irritating to the skin of mice and swine when applied under occlusion for 24 hours. When applied to human skin at a concentration of 2% in petrolatum, it was found to be non irritant following a 48 hour closed-patch test, and was not found to be a sensitiser in 25 human volunteers (Opdyke 1978). Narcissus absolute is prepared by extraction of the flowers with petroleum ether which is evaporated to produce a "concrete", and this is then extracted to produce the "absolute". The absolute contains linalool, benzyl acetate, benzyl alcohol, terpineol, cineol, phenylpropyl and phenylethyl alcohols and their acetates, n-heptanol, n-nonanal, etc.
Aplin (1966) noted that the sap from the cut stems of this species can cause dermatitis. However, he was probably referring to cultivated daffodils rather than to the wild species.
The roots of this and possibly other species are irritant (Burkill 1935).
The juice of Haemanthus multiflorus is supposed to produce dangerous swelling of the lips and tongue (Watt & Breyer-Brandwijk 1962).
[Further information available but not yet included in database]
In the bulb and cut flower industries, those who pick and pack Narcissus species and cultivars are most heavily at risk. It has been estimated that up to 20% of workers may be affected (Schwartz et al. 1957). An expert picker may gather in a day up to 2000 bunches, each of 12 blooms. The skin of the hands and forearms becomes macerated and abraded, and if sensitisation develops, eczematous dermatitis affects mainly the fingers, hands, and forearms, but not uncommonly also affects the thighs and genitalia, and may even be generalised (Walsh 1910, Palmer & Freeman 1934). If warmer weather allows more scanty clothing to be worn, the principal sites affected may be the left forearm and abdominal wall (Marshall 1967).
Bulb handlers may develop dermatitis of the fingers (van der Werff 1959) resembling tulip fingers (see Alstroemeria, fam. Alstroemeriaceae and Tulipa, fam. Liliaceae), but usually less severe.
The dermatitis problem in both the cut flower and bulb producing sections of the industry has on occasion "caused serious dislocation" (Palmer & Freeman 1934).
In florists, dermatitis from Narcissus is less common and is often confined to the hands but may involve the face.
The leaf may be used for patch testing, although tests with stem and flower are also positive (Heyl 1961). Patch tests with the flower and leaf produced positive reactions in a gardener who developed contact dermatitis of the hands from cutting and bunching daffodils. An aqueous extract of the leaves of snowdrops (Galanthus nivalis L.) also produced a positive reaction (Bleumink & Nater 1974).
Although the authors (J.M. & A.R.) have not seen irritant effects from patch tests with leaves, it is a wise precaution always to test some control subjects.