218 Campbell: The embryo-sac of Pandanus 



4. The first division in the embryo-sac results in two similar. 

 nuclei which occupy the poles of the sac. 



5. The micropylar nucleus divides but once; usually no trace 

 of a differentiation into egg-cell and synergid can be seen, but 

 occasionally there seems to be a suggestion of such differentiation^ 



6. The primary chalazal nucleus divides repeatedly until twelve 

 nuclei are formed ; but the order of these nuclear divisions could 

 not be determined, and the twelve resulting nuclei look very much 

 alike. 



7. The nuclei increase in size as the sac grows, but in no cases 

 were there seen any indications of nuclear fusions. In the most 

 advanced stages that were secured, all of the fourteen nuclei were 

 quite separate. 



8. In P, Artocarpus the two micropylar nuclei are somewhat 

 smaller than the chalazal nuclei ; in P. odoratissimiis they are equal 

 in size ; otherwise the two species agree closely in the structure of 

 the embryo-sac. 



Conclusion 



Until the fertilization stages can be examined, it will not be pos- 

 sible to state positively that the fourteen-nucleate stage really rep- 

 resents the condition of the sac at the time of fertilization, and the 

 further history must be left for future investigation. Whether these 

 investigations will reveal in the post-fertilization stages conditions 

 approximating those in Sparganhim, which in its floral structure 

 resembles Pandanus^ remains to be seen. Sparganium in the early 

 development of the embryo-sac follows the'usual course of devel- 

 opment, and it is not until later that there occurs the extraordinary 

 development of the antipodal cells. To judge from the conditions 

 of the sac in the oldest stages that were met with, Pandanus differs 

 very much from the typical angiosperms, and has its nearest ana- 

 logue in Peperomia hispidiila. The increased number of nuclei is 

 evidently perfectly normal, and can hardly have been derived from 

 the ordinary eight-nucleate type, nor can it be explained as a case 

 of fusion of four megaspores, since only one of the megaspore divi- 

 sions is suppressed. Probably the two micropylar nuclei represent a 

 primitive egg-apparatus consisting of the ^g^ and a single synergid, 

 but what relation, if any, the twelve chalazal nuclei bear to the 

 antipodal and polar nuclei of the ordinary embryo-sac it is useless 



