116 DEVELOPMENT OF THE FLOWER 
Then, and lastly, these cells develope amylaceous gra- 
nules. 
Thus it is an oscillation between grumous and cellular 
formations, for even the amylaceous product appears to me 
litle more than a higher development of grumous, both 
consisting of granular matter held together by an apprecia- 
ble, but indefinable medium. 
Iodine only acts in its peculiar way upon the creations 
of the inner membrane ; it acts on all the other granular 
processes in the same way as on ordinary tissue, but in a 
much more intense manner. But of the universal presence 
of this granular matter in all new growths, I am strongly im- 
pressed. Doubtless it is from it that cellular tissue are 
produced. Its disappearance in the young grains as they 
develope, may in some measure be taken as evidence of their 
assisting in the formation of its sulci ? 
Points to be determined. 
Original continuity of polleniferous mass. 
Direction of development to be confirmed. 
Quadrilocularity of anther. 
How the septa are formed. 
Is not the tube a torus.— 
Remarkable formation. 
Ist. Separation or solution of continuity. 
The formation of cells of grumous tissue. 
The formation by these cells of grumous nuclei. 
- The division of these into nucelli. 
Then the disappearance of the granules of these, as the 
grain increases until they are quite observed; then the 
re-development of more distinct ones and in much larger - 
number. Ganges, April, 1841. 
Development of the Pollen, and 2 ovula in 
‘Sisyrinchium. Pi. 
When the perianth is 1 line long, pé anthers are highly 
developed ; but they are Me Each cell, of which there 
* The plate in lithographing was erroneously entitled Pardanthus Chinensis. 
& 
