﻿BOTANICAL GAZETTE 



By the end of September the flower buds are well formed and 

 the parts of the ovules are clearly differentiated (fig. i). An arche- 

 sporial cell immediately below the epidermis has divided and given 

 rise to a primary parietal and a megaspore mother cell (fig. 2). In 

 some ovules the primary parietal cell remains undivided (fig. 2), 

 while in others it divides longitudinally (fig. 1). Ernst (loc. cit.) 

 finds no formation of primary parietal tissue in T. grandifiorum. 



During the winter, the megaspore mother cell of T. cernuum 

 is in a resting state and is easily distinguishable from the surround- 

 ing cells because of its larger size and its larger and more deeply 

 staining nucleus (figs. 1,2). If meiosis is accepted as the criterion 

 for spore formation, the fact that this cell later undergoes the reduc- 

 tion divisions proves it to be a megaspore mother cell. The 

 physiologically analogous cell of T. grandifiorum is called by Ernst 

 (loc. cit.) the embryo sac mother cell. He says: "Die Embryo- 

 sackmutterzelle differenzirt sich in der subepidermalen Zellschicht 

 unmittelbar unten dem Scheitel des Nucellus." Since, as in T. 

 cernuum, this cell does not give rise directly to the embryo sac, it 

 cannot be the true embryo sac mother cell, and this terminology 

 should be discarded. 



With the resumption of growth in late March or early April, 

 active cell division may be observed in all parts of the flower. The 

 resting reticulum of the megaspore mother cell gradually loses its 

 netlike structure and resolves itself into a dense synaptic mass 

 (fig. 3). Later stages show the chromatin threads to have thick- 

 ened, shortened, and segmented transversely into distinct chromo- 

 somes (fig. 4). After a second contraction of the nuclear material 

 (fig. 5) the separate chromosomes soon become arranged at the 

 nuclear plate, while delicate spindle fibers can be distinguished in the 

 cytoplasm (fig. 6). 



Anaphase and telophase are quickly passed through, as Ernst 

 (loc. cit.) has reported for T. grandifiorum, and each of the resulting 

 daughter nuclei becomes invested with a delicate nuclear mem- 

 brane. The chromosomes of each nucleus seem to fuse end to end 

 to form a long, thick, loosely wound band, but there is no evidence 

 from the material studied that they completely lose their identity 

 at this time and form a reticulum. During this short resting 



