sporiiom is round abo\it thf? rnidcllc of June. The sub- 

 epidei'inal layer divides at the place where the two pol- 

 len sacs are to be fox-med by periclinal v/alls.(Fic. 4) 

 The exact derivation of layers is hard to trace, but 

 it is quite coi'tain that it is from the inner layer 

 thus formed that all the archesporial tissue comes. By 

 the middle of July the archesporium is well blocked out 

 and shortly after the spore mother-cells are formed. At 

 this stage there is about these a moderately well-de- 

 fined layer of tapetum, and the outside wall of the 

 pollen-sac is three or four cell layers in thickness 

 (Fig. 5). Here the pollen mother-cells are only notice- 

 able by their slightly larger nuclei and more deeply 

 staining contents. The further growth of the microspor- 

 angium is brought about by the increase in size of its 

 cells, botlt archesporium and tapet\im. 3efoi-e the tet- 

 rad division the nuclei of the tapetum divide without 



