To a science which prides itself upon precision of language, any confusion in terms is a reproach. The terms of systematic botany enjoy a respectable antiquity, but sometimes lend a regrettable obscurity to the scene. The names applied to inflorescences were confused from the beginning. In current manuals they are usually classified as determinate (cymose) or indeterminate (racemose). Umbels and panicles are included among the latter; yet the same manuals will refer to the compound determinate clusters of Allium, Pelargonium and Asclepias as umbels and to the aggregated cymose clusters of Gilia and Penstemon as panicles. The inflorescence of Claytonia is commonly described as a raceme, whereas to a more than casual glance it is evidently cymose. Usage of botanical terms can not be settled, like that of botanical names, by a principle of priority; it would not advantage the science to compel taxonomists to adhere to the definitions of Linnaeus. An alternative is to adopt current usage; in view of the somewhat chaotic conditions indicated above, this also seems an unreliable guide. To reverse the old dictum: Res si nescis, perit et cognitio nominum. A rational terminology mirrors that upon which it is based-an understanding of the things concerned. To discuss the terms which have been applied to inflorescences (or to flowers, or to fruits) is to examine the concepts through which our current understanding (or lack of understanding) of these parts has been developed. Such a survey may indicate the problems that await investigation, besides clearing away the accumulated rubbish of the ages. In short, an attempt to arrive at a rational terminology is ipso facto an effort to evaluate scientific concepts and generalizations. Much knowledge of inflorescences has accumulated since the first great flowering of botany in the eighteenth century; and many theories. Unfortunately, names and classes of inflorescences have been proposed under the influence of certain theories now discredited or which ought to be discredited. and in the absence of 187