— 156 — 



Meanwhile it was soon evident, that not all ovules had EMC 

 with a typical tetrad division. In some rare instances I have found 

 that something like the case which Juel has described for Tarax- 

 acum, has taken place here, i. e. there is only one division of the 

 EMC and with an unreduced number of chromosomes. I found 

 spindle figures from the first division in which the number 01 

 chromosomes was clearly greater than the reduced one and pro- 

 bably like the somatic one. In the fig. 19 and 20 the upper cell 

 represents the embryo sac cell, which has already begun to encroach 

 upon the under cell, and its nucleus is evidently in the resting 

 stage. Such cases, however, are so rare that they cannot explain 

 the great number of apogamic ovules that Ostenfeld (16) dis- 

 covered. 



Moreover I found, that in many ovules, where the EMC had 

 been divided into tetrads, another simultaneous process occurred. A 

 somatic cell close to the EMC become extended and began to 

 show every sign of being an embryo sac. Often this cell was 

 situated behind the tetrad-row as in fig. 25, but also further away 

 in the chalaza-region. It gradually encroaches, however, upon 

 the cells of the tetrad, fills up the nucellus and becomes an 

 embryo sac. 



The nucleus of this cell divides in the same manner as in an 

 ordinary embryo sac : there are formed 3 antipodal cells, 3 syn- 

 ergids, 1 egg cell and 2 polar nuclei which move toward one another 

 and fuse, forming the primary endosperm nucleus. In the afore- 

 mentioned note I have compared this embryosac forma- 

 tion with the apospory in certain ferns. 



The number of chromosomes is, of course, the unreduced one 

 which I have been able to observe in several cases. It should be 

 observed, that this somatic cell does not directly develop into an 

 embryo, but first forms' an ES, i. e. a gamophyt generation, in 

 which the egg cell without fertilization develops into an embryo. 

 In cases such as are represented by fig. 25 and 26 one could 

 possibly be doubtful if the great cell was not really the typical 

 ES in spite of the fact, that besides there is a tetrad; the one cell 

 of the tetrad could perhaps represent a sort of parietal cell. Such 

 cells, however, are not found in Hieracium and fig. 27 shows 

 clearly that the cell Ap is quite another cell outside the tetrad. 

 It is interesting to observe that in the tetrad cells in the fig. 27 

 dwarfed nuclei can be traced here and there, which proves 



