CALLUNA, CASSIOPE AND HARRIMANELLA: A TAXONOMIC AND EVOLUTIONARY PROBLEM

Leaf types in Calluna, Cassiope and Harrimanella are described; the so-called needle leaf of Cassiope lycopodioides and C. mertensiana is shown to be derived from the channelled leaves found in the majority of species of Cassiope, possibly via an intermediate stage such as C. redowskii, which has leaves like the finger of a glove. The needle leaves of C. palpebrata are superficially different from the other leaf types of the genus, as are those of Harrimanella. A considerable number of differences justifying the maintenance of Cassiope and Harrimanella as separate genera are listed. The relationship of Cassiope and Harrimanella to the Rhododen-droideae and Vaccinioideae are discussed, and although these two genera are very isolated in the Ericaceae they are closer to the Andromedeae than to any other group. With Calluna they have little in common apart from characters occurring sporadically throughout the family; although Cassiope and Calluna are similar in a number of points Harrimanella and Cassiope differ in these characters. The evidence suggests that Calluna is a member of the Ericoideae, although separable as a monotypic tribe; Cassiope and Harrimanella are not at all close to the Ericoideae. It is concluded that leaf type is a character which must be used with caution in establishing relationships within the Ericaceae; the evolutionary significance of the rather frequent conflict between relationships suggested by leaf type and those suggested by floral characters is briefly discussed.