48 



BOTANICAL GAZETTE 



[JULY 



seems ordinarily to be far too large. Such a wall, in contact with the 



fig 



This wall persists 



around the tetrad until late in the development of the spores. The 



asm 



The 



vacuole is limited by a plasmatic membrane often more prominent 

 than the wall of the mother cell within it. Where the two are in con- 

 tact they are distinct, showing that the vacuole is a real one and is 



erial. I was 



ma 



not able to determine just how long the old wall persists, but it was 

 found (at least parts of it) around spores as old as those shown in 



fig^ 27 ' 



The beginning of the tetrad walls is shown in fig. 21 and the com- 



fig 



primary 



spindles of the second division and partly on secondary spindle-like 



structures 



Of 



the six walls which meet in the middle and separate the young spores, 

 two form on the primary spindles and four on the secondary ones. 

 At the same time a wall forms on the outer or convex side of the 

 spores. Whether this wall joins directly to the tetrad wall or con- 

 tinues around inside of it could not be determined. Beer describes 

 this tetrad wall as uniting with the wall which surrounds the tetrad, 

 thus forming a structure with four compartments, each of which con- 

 tains a spore. It is certain that it does not do so in Ophioglossum, 

 owing to the fact that the mother-cell wall is not at that time in contact 

 with the spores. 



spores 



The 



resting nucleus has a rather coarse network, with the chromatin 



masses (fig 



vacuolate 



As the spores grow 

 cytoplasm becomes 

 oulasm seems to be 





.ponding 



coats. 



ght out in figs. 22-26. 

 During this period of thickening of the spore walls the cytoplasm 



vacuolate. A decrease 

 Lcuolation of the plas- 



more 



pari p 



modium. 



The degree of the collapsing action of the killing and 

 ts on the young spores affords a graphic though not very 



- i 



cv 



V 



- ^ 



*' 



