What KIAERSK wrote in 1893 in the preface of his “Enumeratio Myrtacearum Brasiliensium” is still largely valid. It is often most difficult to define a species belonging to this family, not only because, in the absence of ripe seeds, the genus is not easily ascertainable, but also because of the strong variability shown by the vegetative characters. Thanks to the examination of the rich Guiana material preserved in the herbaria of Genève, Kew, Leiden, New York, Paris and Utrecht, I have usually been able to delimit the species in a satisfactory way; their allocation to a definite genus, however, is often a difficult problem. During the preliminary stage of this investigation, which was interrupted by the war, it was of great advantage to me that I could study the Guiana specimens of the Leiden herbarium. In order to avoid misinterpretations, I have tried to base my conclusions as far as possible on an examination of either the types themselves or of duplicates of the latter. Several of these types, especially those that form part of the earlier collections of Guiana plants, e.g. of the collection Aublet, and of the collections Desfontaines (herb. Florence) and De Candolle (Genève) had never before been reexamined, and BERG, the last monographer of the South American Myrtaceae (in Linnaea XXVII (1855—56), XXIX (1858) and XXX (1861) has either neglected these species or given an, often incorrect, interpretation based on the description alone. For this reason the second part of this paper will be devoted to a short survey of these earlier types. My best thanks are due to the directors of all herbaria mentioned. Moreover, I have to thank the “Van Eedenfonds”, whose financial aid enabled me to pay a visit to Kew and to the British Museum.