48 BOTANICAL GAZETTE [ JANUARY 
‘ 
youngest anthers in which archesporial tissue could be detected, 
no indication of the exact region to become sterilized could be 
made out, since all the cells stain alike. However, one may 
soon distinguish the larger, more lightly stained, and less rapidly 
dividing cells which are to form the sterile plate. In Mazs 
fiexilis Campbell (14) observed that but one archesporial mass 
is formed in each anther; that this multiplies its cells for a con- 
siderable time before there occurs any differentiation into tapetal 
and sporogenous regions; and that ‘“‘even in later stages the 
boundary between the sporogenous cells and those lying outside 
is not always perfectly clear.” But in this mass of archesporial 
tissue no sterilized separating regions were formed, the entire 
mass forming a unilocular sporangium. 
Such conditions suggest certain pteridophytes, where the spo- _ 
rangia are developed in a manner not unlike those of L. nor. 
In Isoetes we have an archesporial mass which, after having 
grown until it consists of a large number of cells, develops 
plates of sterile cells, the trabecule, which separate groups of 
sporogenous cells more or less completely from one another. In 
Isoetes the number and arrangement of these groups of sporog- 
enous cells are irregular, while in LZ. minor they are regular; 
but in other respects the two present many points in commoi. 
The origin of the archesporial tissue seems to be the same, 
since in each it originates as a hypodermal layer; in behavior 
it is essentially the same, since in both it divides rapidly, 
forming a mass of cells, some of which become sterilized as 
plates which separate groups of sporogenous cells; and in 
each the tapetum is formed after these sterilized plates are com- 
pleted. 
If my observations and interpretations are correct, and if the 
spore bearing region of Isoetes is a single sporangium, the four 
groups of sporogenous cells of one anther of L. minor must be 
the same. Whether in angiosperms other than L. minor it can 
be shown that each loculus of the anther is not a real sp? 
rangium, as given in current accounts, but rather that two OF four 
loculi of one anther are together the sporangium, must P 
