263 



Amaryllidaceae. 



Very scanty information seems to be had about microsporogene- 

 sis in Amaryllidaceae. For all till now investigated plants, viz. 



Agapanlhiis 



Cnrculujo recuruata Jack. 



Eucharis 



Hgmenocallis litoralis Salisb. and 



Po lian t hes tuberosa, 

 successiv bipartition was recorded. Generally a conspicuous cell plate 

 is laid down in the helerotypic spindle. Growing centrifugally, it 

 becomes attached to the pollen mother cell membrane where 

 this membrane forms a slightly projecting equatorial belt. The 

 second partition is brought about in the same way. 



Taccaceae. 



In the Liliiflorae simullanous (juadripartition has been observed 

 in the following families: Dioscoreaceae, Iridaceae, Jnncaceae and 

 Liliaceae. To these the family of Taccaceae might now be added 



Tacca cristata Jack. 

 showing the same mode of pollen development. 



In several preparalions, however, young tetrads were met with 

 where the four defmitive cells were forming a row, indicating thus 

 the possibility that successiv bipartition had occurred. The teträds 

 seen, reminded slrongly of those, figured by Levink for Drosera. 

 Further work is needed to seltle this question. 



Bromeliaceae. 



An investigation of two species of this family, 

 Crijpthanthus acaulis Beer and 

 Pitcairnia spec. 

 established the successiv bipartition for bolh of them. 



Commelinaceae. 



Members of this family have repeatedly been investigated; In all 

 recent publications successiv bipartition seems to have been re- 

 corded; nor do. 



