1899] DEVELOPMENT OF THE MICROSPORANGIUM 355 
SUMMARY. 
The following brief summary may aid in bringing together 
the results reached in the foregoing pages. 
The experiments with regard to the effect of external condi- 
tions on nuclear division both in Convallaria and Potamogeton 
gave no results for light and humidity, which were the only con- 
ditions tested. 
The material illustrating the younger stages in the develop- 
ment of the microsporangium shows that the process is slightly 
different in Convallaria and Potamogeton from the normal 
method as given by Warming and Engler. The archesporial 
tells arise by the division of a hypodermal cell at one corner of 
the anther. Therefore, instead of the archesporium arising from 
alayer of hypodermal cells, as Warming describes for dicotyle- 
dons, it arises from one or rarely two hypodermal cells. The 
primary archesporial cells divide only a few times, but there is 
‘siderable subsequent growth in size of each cell. The next 
outer cell in the original row forms part of the tapetum, and the 
tmainder are wall cells. Most of the wall and tapetum, how- 
"Wer, is formed from the tissue at either side of the archesporium 
| “amine seem to show that this is more likely the normal pro- 
Die for the whole group. 
: The tapetal nuclei of Convallaria show nicely the process of 
pear fusion which has been described by Strasburger and 
: for many other plants. After the division of the primary 
a nucleus by the mitotic method, the two daughter nucle! 
Fi many cases fuse again, and all stages of the process may be 
: a piten in the same anther. It is probable that not all the 
% divide, and also that not all of those that do divide fuse 
“ae before disintegration. It seems that in Potamogeton no 
pet the tapetal nuclei takes place. 
© structure of the wall in the mature microsporangium 
