HÉILBORN, CYTOLOGY OF ANANAS SAT1VÜS LIKDL. 3 



he number of univalents at 14—18. In the following divi- 

 ions the univalents are irregularly distributed, dwarf nuclei 

 .rise, the tetrads thus formed presenting a very abnormal 

 jppearance (Fig. 3). The plasma very soon becomes vacuo- 

 ated and the pollen-grains collapse and degenerate completely. 



€L. 



Fig. 1—2. Heterotypic raetaphasea in pollen mothcr-cell.s of the Banos- 

 race. — Keich. horn. imm. V", Leitz peripl. ok. 20 X. 



In the tapetal cells of very young anthers — the mother- 

 cells not yet having been loosened and rounded off, and the 

 tapetum not yet showing any degeneration phenomena — 

 the somatic chromosome number could be counted in several 

 metaphase plates. The tapetal cells are rather large, and 

 the somatic chromo.somes small and almost spherical, so these 

 metaphases appear with remarkable clearness (Figs. 5—7). 

 Only the very best plates were selected and, in five out of 



