SIGNIFICANCE OF DICHOGAMY. 317 



ihere are many whose flowers are not protogynous but protandrous. Here the 

 stigmas of the earliest flowers of a species cannot be pollinated, because they are 

 .mmature and inaccessible. What, then, becomes of the pollen of these first prot- 

 indrous flowers ? If it is carried by the wind or by insects, as soon as it is liberated 

 :rom the anthers, to any stigma, that stigma must of necessity belong to another 

 species which has already become receptive. Towards the end of the flowering 

 period, the pollen usually runs short in most protandrous species, whilst the stign 

 rf these stragglers have only just attained maturity. They could only obtain 

 pollen from flowers which had not developed so far. But if these flowers are the 

 last in the locality, and they are protandrous, there is no more pollen to be had 

 from that species, and obviously they must be satisfied with pollen from some other. 

 Accordingly hybridization is a matter of necessity in the latest flowers of herma- 

 phrodite plants which are protandrous, just as it is in the earliest flowers of those 

 which are protogynous. 



From these facts we may infer that every dichogamous plant has an opportunity 

 for illegitimate crossing or hybridization at the beginning or end of its flower- 

 ing, and that dichogamy especially incomplete dichogamy is the most important 

 factor in its production. Of course this does not exclude dichogamy from playing 

 an important part in legitimate crossing as well. On the whole, however, we can 

 maintain the view that the separation of the sexes by the maturation of the sexual 

 organs at different times leads to hybridization, whilst their separation in space 

 promotes legitimate crossing. The fact that the separation of the sexes in time 

 and space usually occur in conjunction, harmonizes with this conclusion, i.e. that 

 the dioecious, monoecious, and pseudo-hermaphrodite flowers, as well as those herma- 

 phrodite flowers whose sexual organs are separated by some little distance, are 

 in addition incompletely dichogamous, because by this contrivance the flowers of 

 any species obtain (1) the possibility of hybridization at the beginning or end of 1 

 flowering period, and (2) of legitimate crossing during the rest of that time. 

 exDlains why incomplete dichogamy is so much more frequent than complete dicho- 

 gamy why there are no dioecious species of plants with completely dichogamoi 

 flowers- and why, if one ever should occur, it would of necessity soon disappear. Let 

 us suppose that somewhere or other there grows a species of Willow with complc 

 protogynous dioecious flowers, that is to say, a species in which the female 

 mature first, and have ceased to be receptive before the male flowers m the same 

 region discharge their pollen. Hybridization only could occur m it, and 

 fallow plants resulting from it would all be hybrids whose form wou d no longe 

 agree absolutely with that of the pistilliferous plant. The species would ther 

 not be able to reproduce its own kind by its seed, and it would leave no descen- 

 dants of similar form ; in other words, it would die out. 



