THEORETICAL CONCLUSIONS 417 



tet, i.e., the antipodal cells and lower polar nucleus, show a great 

 variation in their behavior which is quite unknown for archegonia. 

 Usually the antipodal cells are ephemeral and may disorganize 

 even before any wall formation has taken place between the nuclei. 

 In the Podostomaceae the primary chalazal nucleus remains un- 

 divided, and in the Oenotheraceae there is no nucleus at the chala- 

 zal end. These must be considered as instances of a tendency 

 towards the reduction and final elimination of the chalazal arche- 

 gonium. There are other plants, however, in which the antipodal 

 cells persist and become very active. Sometimes they show nu- 

 clear divisions inside them, followed by fusions resulting in a high 

 degree of polyploidy. In other cases, the divisions are accompanied 

 by wall formation, resulting in a massive tissue which persists for 

 a long time. Finally, in several plants, like Drusa, Tanacetum, 

 and Chrysanthemum, even the initial number of antipodal cells ex- 

 ceeds three, and then the so-called archegonial plan cannot be 

 recognized at all. 



6. Some insuperable difficulties arise in applying Porsch's inter- 

 pretation to the bisporic and tetrasporic embryo sacs. In the bi- 

 sporic sacs the micropylar archegonium is derived from one mega- 

 spore nucleus and must correspond to one prothallus, while the 

 chalazal archegonium is derived from another megaspore nucleus 

 and must therefore correspond to a second prothallus, i.e., the em- 

 bryo sac is composed of two prothalli. Proceeding on the same 

 analogies, in tetrasporic forms, like Adoxa, each archegonium must 

 be supposed to represent two prothalli, for the synergids (i.e., neck 

 cells) are derived from one megaspore nucleus, and the egg and 

 upper polar nucleus (i.e., ventral canal nucleus) from a second 

 megaspore nucleus. In the Fritillaria type, the micropylar arche- 

 gonium represents one prothallus, but the chalazal will have to be 

 considered as the equivalent of three prothalli which fuse at the 

 megaspore stage. Strangest of all would be certain forms of 

 Tulipa, belonging to the section Eriostemones, for here the micro- 

 pylar archegonium, which consists of more than three cells, must 

 be regarded as having originated from three megaspores. 



That a single archegonium should correspond to one, two, and 

 even three prothalli is incomprehensible, and it seems that the 

 very simplicity of Porsch's theory, which led to its adoption in the 

 past, must now be the ground for its final rejection (see also 



