FERTILIZATION 301 



reported method is thai by which the two polars fuse to form an "em- 

 bryo sac nucleus'' before the entrance of the pollen tube, the male 

 nucleus later being added. Ernst (1902), for example, found this to I 

 the method in Paris quadrifolia. I.<ss frequently the male nucleus 

 meets and fuses with the polar nucleus of the micropylar end of the sac, 

 the other polar then fusing with the product. This is the method de- 

 scribed by Nawaschin (1898, 1899) in his accounl of the discover} 

 double fertilization in Lilium Martagon and Fritillaria ienella. The 

 simultaneous fusion of all three nuclei appears to 1"' a common 

 occurrence; it has recently been described in some detail by Miss Nbth- 

 nagel (1918) for Trillium and Lilium. Just as in the case of the union 

 of the first male nucleus with the egg nucleus, the chromatin of tin- 

 second male and two polar nuclei may be either in the reticulate or the 

 spireme condition as they come together. In Lilium Martagon Nbth- 

 nagel 1918) (Fig. 123, D) it is in the form of fine strands, intermediate 

 between the resting and spireme stages. Although the three nuclei often 

 appear exactly alike, it is frequently possible to distinguish the male 

 from the polars, not only by its shape and -mailer size, but by the 

 condition of its chromatin: in Lilium longiflorum (Weniger 1918 . 

 example, the male nucleus is in the spireme stage while the polar nuclei 

 are still in the resting condition. The membranes of the three nuclei 

 may persist for some time after they come into intimate contact, and 

 even after they have dissolved the chromatic element- of the three con- 

 stituent nuclei may in many cases be distinguished it' thi tion has 

 been made in a favorable plane. When fusion occurs in the resting 

 stage this is not so apparent, but when it occurs in the spiren 

 the three chromatic groups are made out with little difficulty. 



Endosperm. — As the division of the endosperm nucleus approach 

 the spiremes of its three constituent nuclei become increasingly dis- 

 tinct, even if one or more of the nuclei have fused in the resting Btaf 

 Nothnagel (1918), Weniger (1918), and Sax (1918) in their recent Btudies 

 all report this condition (Fig. 123, F). As the spiremes are beii 

 developed into completed chromosomes all of them :;\ in number Bplif 

 longitudinally, no observer reporting such a pairing as some have thought 

 to occur between the chromosomes of the egg and first male nuclei. 

 Miss Nothnagel describes the formation of a tripolar spindle about the 

 chromosomes; the bipolar condition soon develops from this. Hon 

 frequently this may occur is not known. Eventually in any case the 

 mitosis proceeds along the usual lines and the two daughter nuclei n 

 3x chromosomes each. This number is characteristic of all the cells of 

 the endosperm formed by the repeated division of these nuclei. An 

 exceptional condition has been noted by Sax L918 in FriHUat 

 Here the lower polar nucleus, because of an irregularity in the mitosis 

 giving rise to it, has 21 (2x) chromosomes instead of the normal 12 



