.118 DEVELOPMENT OF THE FLOWER 
open along the centre throughout the whole of their outer 
faces, they are papillose and reticulate ; they contain a 
hyaline sac, apparently empty, which ii out somewhat ` 
through the chink mentioned above. Fig. 
Perianth 3 lines.long. The mass remains AA the same 
exteriorly, the envelope and grumous matter stil being 
present, but this last is confined to the immediate neighbour- ' 
. hood of the envelope. The grains of pollen are now nearly 
perfectly formed. They vary much in size, are minutely 
reticulate, opaquish, and marked along one side with a trans- - 
parent line—the remains of the chink mentioned above. 
They are but little acted 'on by protracted immersion, no- 
traces of the originally enclosing cells are visible. Fig, 6. 
Perianth 5lines long. The grains remain of the same ap- 
pearance, and are excessively unequal in size, the smaller ones 
being probably? abortive. "The chink is nearly closed by 
the rolling in of the margins. They are now susceptible to 
the action of water: the lips of the chink expanding until at 
_ length they are widely separate. Theliving membrane is very 
distinct and bulges out considerably after protracted immer- _ 
sion in water, but it has no tendency to separate. Fig. 7, 8. 
Perianth 6} lines long. Pollen perfectly formed, varying 
much in size, but all the grains undergoing the same change. 
on immersion in water. The outer membrane is rapidly . 
pushed back on immersion by the enlargement of the inner 
hyaline one, which is now nearly filled with granules, still it 
does not spontaneously separate, although os traction or 
pressure is sufficient to insure this. Fig. 9 
At the time of expansion the granules vary in size, the most 
` perfect being lanceolate. They are highly reticulate, and the 
chink is very visible. They are of an orange-yellow colour. 
Immersion in water causes a sudden bulging out of the 
inner membrane, which at length pushes back, and finally 
off, the outer coat. It is filled with granules and soon nares 
the granules are active molicules. Fig. 10. 
This is interesting as an instance of development of pollen 
from the ternary or quaternary division of an originally sim- 
ple nucleus, as an instance of pains and almost page 
* : * * 
