402 BOTANICAL GAZETTE [December 



seems to be the most satisfactory hypothesis at present. Since 

 no case of retardation in division of the second daughter 

 cell was observed, it would be interesting to know the relative 

 time of division of the two when the first or second megaspore 

 predominates. Since only one case of the division of the daugh- 

 ter cells was observed in A. tuberosa, no conclusions can be drawn 

 concerning the meaning of the smaller spindle in the first. 



The embryo sac passes rapidly through the typical two, four, 

 and eight-nucleate stages {^figs. 42, ^j), and the usual ante-fertil- 

 ization stage of the sac is reached. Occasionally one finds more 

 than three antipodals {^fig. 45), but it is quite the exception, and 

 only once were fewer than three noted. The synergids resem- 

 ble the egg very closely in many cases and can be distinguished 

 from it only by the relative nearness of their nuclei to the micro- 

 Py^^ {.fiS^'43~4^) ' The number of cells at the micropylar end 

 was constant, although Chauveaud (3) figures five micropylar 

 cells in'the sac of Cynanchum,and believed that the five embryos 

 sometimes found were due to the fertilization of all of them. 

 The sac enlarges very rapidly, soon destroying the nucellus and 

 much of the surrounding tissue. In one case noted it had 

 replaced the surface of the ovule {fig- 44) ^ while in several 

 instances it had come near it. 



In this connection it may be noted that in two ovules a tra- 

 cheid was observed near the base of the embryo sac and slightly 

 projecting into it, showing that the thickening of their walls had 

 occurred before the sac had reached its full development {fig* 

 48). In only two other genera of angiosperms has the occur- 

 .rence of tracheids in the nucellus been reported : in Castanea by 



ss Benson (i), and in Casuarina by Treub (15)- The occur- 

 rence of a tracheid near to such a rapidly enlarging embryo sac 

 raises the question whether increased conduction had not been 

 the cause of its formation. That increased conduction leads to 

 increased formation of tracheary tissue seems evident from Voch- 

 ting's experiments with the potato and other tubers. That it 

 causes the formation of tracheary tissue where there has been 

 none phylogenetically there is no evidence except in the evolu- 

 tion of vascular from non-vascular plants. Could the tracheids 



M 



