Nov., 1905.] Life History Saiiguinaria canadensis. 381 



of the microspore had enlarged so that the nucleus was forced 

 to one side of the cell. Material taken on March 13th showed 

 but little change from that taken on the previotis date. The 

 megasporocyte had enlarged quite a little and seemed to be pre- 

 paring to divide (Fig. 4), and a few spindles were observed 

 (Fig. 6). The microspores showed no change. 



Material fixed on March 22nd showed that the megasporocyte 

 had divided into four megaspores the lowest being the large 

 functional one (Fig. 7). The arrangement of the megaspores 

 varied considerably. In some cases the transverse walls were 

 nearlv parallel forming a row of cells but frequentlv they were 

 verv irregular (Figs. 7, 8). These divisions are interesting in 

 that the\' show a great similarity to the divisions of the micro- 

 sporocyte and often result in a nearly typical tetrad (Fig. 8). 

 Above the megaspores the rows of parietal cells could be dis- 

 tinctly seen (Fig. 7). These varied in number but it is evident 

 that considerable division has occurred in the three original 

 parietal cells. Vesque (1) states that Papaver orientale has no 

 parietal cells. The nuclei of the microspores were divided at 

 this time (Fig. 21). Pollen grains taken at later dates did not 

 seem to show any further division. It is probable that the gen- 

 erative cell does not divide until after the tube has been formed. 

 Strasburger (2) however, reports that in Papaver the generative 

 nucleus divides in the pollen grain. 



From this time the development of the functional megaspore 

 is rather rapid. Material taken on March 28th showed the com- 

 plete eight celled embryo-sac and often the two polar nuclei had 

 already fused to form the definitive nucleus (Fig. 9). The 

 antipotals are rather large at this stage and each has a single 

 nucleus. Material taken on April lOth showed the oosphere 

 and definitive nucleus still undivided but the antipodals were 

 verv large and each contained two nuclei (Fig. 10). This stage 

 corresponded with material taken the vear before at al)Out the 

 same date. The egg then seems to be fertilized about the first 

 week in April. Many of the sections show remains of the pollen 

 tube which is very prominent (Figs. 10, 11). The svnergids 

 seem to disappear early. Remains of one or more of them could 

 usuallv be seen at this stage, Iving to one side of the oospore 

 and staining very dark (Figs. 10, 11). The earlv divisions of 

 the embryo seem to occur very slowly for it was not until May 

 12th that the first division was observed (Fig. 13). This makes 

 a remarkablv long period of rest for the fertilized egg. In 

 material killed May Kith the three celled eml)ryo was found 

 (Fig. 14). The divisions of the definitive nucleus had l)cgun in 

 earlv April and bv the latter part of that month the en^j^f^^jn^ 

 had' 20 to 30 nuclei. 



