FACTORS OF DECREASE 127 
Arisarum, and in Najas, §Caulinia, can hardly be ascribed to any other 
source than the fusion of the pollen-sacs, separate in the ancestry, into a 
single loculus. Thus in the androecium of Angiosperms, and occasionally 
Juniperus communis. 1.,summit of a male flower seen from above. st), the uppermost 
staminal whorl of these stamens ; s¢j, the second staminal whorl shows on each stamen 
two pollen-sacs, and the indication of a lamina, 2; s¢s, the third staminal whorl, of which 
only the tips of the lamina: of two stamens are seen : each of the stamens of this whorl 
had three pollen-sacs, not shown in the figure. II., the same in longitudinal section. 
III., the same in transverse section. There is evidence here of reduction of the lamina, 
and of fusion of the pollen-sacs. (After Goebel.) 
in the gynoecium, a fusion of sporangia is recognised, resulting in a reduc- 
tion in their number. The apex of the male flower in Juniperus communis 
has been quoted by Goebel as a probable example of fusion of pollen-sacs 
(Fig. 69). In certain Pteridophytes the grouping of 
the sporangia is often such as to suggest a previous 
fusion ; but this has not been proved on developmental 
or comparative grounds for any one case, and the 
question must be left open for discussion on grounds 
of general probability whether the synangial state in 
any individual case has been the result of fusion, or of 
septation with incomplete separation of the sporangia : 
obviously the synangial structure would be compatible 
with either origin. Whatever the final decision for the 
Pteridophyta may be, it is clear that fusion of sporangia 
originally separate has actually occurred in Flowering Fic. 70. 
Plants: it is therefore a factor which must be regarded Saenel Bt aneniene 
A ; a : emorosa, 1. Showing 
as a possible explanation of all synangial states. Beene develaped, aiid 
ae ‘ 3 ioe : apa a three abortive. (From 
(g¢) Abortion, partial or complete, of sporangia which rer ty Prantl 
were fully matured in the type or ancestry is so common 
a feature that special examples need hardly be quoted. Abortive pollen- 
sacs are commonly found on staminodes, and abortive ovules are frequently 
seen, as in Anemone (Fig. 70), which can only be accepted as the imperfect 
representatives of a plurality of ovules in the ancestry. In the Pteridophytes 
many examples of abortive sporangia have lately been described, and their 
very important morphological bearings will be discussed at length in 
Chapter XIII. But in connection with the circumstance that parts 
