In a recent analysis of all the available testimony in reference to the morphological nature of the endosperm of angiosperms, it seemed clear that certain conclusions might be reached, and the purpose of the present paper is to state them. It has been assumed that the endosperm must be either gametophytic tissue or sporophytic tissue, and the arguments for each view are familiar. The perplexity has arisen chiefly from the feeling that gametophyte and sporophyte must be subject to rigid definition. When definitions become rigid, ideas become rigid also, and nature is always playing havoc with rigidity. If gametophytes and sporophytes are defined as x and 2X structures, respectively, straightway x sporophytes and 2X gametophytes are discovered. If sporophytes are defined as structures produced by fertilized eggs, the definition is contradicted by numerous sporophytes that are not the product of fertilization. In this way, every criterion suggested has found its contradiction. It is becoming evident that definitions must be elastic and not rigid, and that general situations rather than definite categories must determine conclusions. We have moved so far beyond the rigid categories of the days of metamorphosis, that it is surprising to find an equal rigidity in the days of alternation of generations. Without citing an extensive and familiar literature, attention may be called to the various claims that have been made as to the morphological nature of the endosperm of angiosperms. Ever since the comparative morphology of the vascular groups was uncovered by HOFMEISTER, belief has been general that the endosperm of angiosperms is gametophytic tissue which develops after fertilization. It was easy, even in the days of HOFMEISTER, and much more so now, to obtain from gymnosperms what seems to be abundant confirmation of this claim. Throughout that group there is a distinct tendency to differentiate eggs earlier and earlier Botanical Gazette, vol. 51] [380