STUDIES IN THE DEVELOPMENT OF THE PIPERACEAE 719 
sort were shown by Engler ('76, p. 297), to occur in the Asclepia- 
daceae, and, in Asclepias syriaca, where the two posterior spo- 
rangia are wanting, this absence is due to the complete failure 
of the archesporium to appear even at the earliest stages of devel- 
opment. 
The second sort of reduction mentioned is clearly the kind re- 
ferred to by Goebel as a '^confluence of pollen sacs," which proc- 
ess, he says (p. 554) 'may take place by the subsequent break- 
ing down of sterile tissue, or b}^ the development of fertile tissue 
in places where otherwise sterile tissue should be." Bower refers 
to this method of decrease in the number of sporangia as the 
^'fusion of sporangia." 
It is to be noted that this decrease in the number of sporoge- 
sous masses may not mean a decrease in the number of spores 
formed, if the size of the persisting spore masses is sufficiently in- 
creased. This is evident from a comparison of figs. 12 and 15, or 
20, 21, 24 and 26, all of which are from stamens of about the same 
age and are magnified to the same degree. 
The important questions to be answered concerning these 
peculiar stamens of Piper betel are : 1. At what stage of develop- 
ment is the suppression of one, two, three or all four of the micro- 
sporangia normally present, first discoverable? 2. At what stage 
is the body of sporogenous cells, that in certain stamens seems 
the topographic equivalent of two or three normal sporangia, 
first recognizable as a continuous mass? 3. In what manner are 
these peculiar modifications of structure brought about? In 
other words, to put all three questions in one, at what time, in 
what manner, and from what cause, does the trend of develop- 
ment in the various tissues of the anther depart from the usual 
development of these in the normal stamen? 
The shape of the upper pollen sacs in the stamens shown in 
figs. 23, 24, 27, suggests that the continuous sporogenous mass 
there shown is a result of the partial disappearance of a septum, 
which at an earlier stage separated two distinct archesporial 
groups in the upper theca. The same conclusion might sometimes 
seem warranted by the appearance of sections of the same stamen 
taken at different levels (figs. 22, 23) . The former of these figures 
