Petrie. — On the Pollination of the Puriri. 411 



placed and kept for a considerable time in the axis of the corolla- 

 tube are evidently designed to bring about pollination of the 

 pistil by the pollen of some other flower. That other flower 

 may be on the same tree or may be on another. Whether the 

 pollen from another tree is prepotent over that of other flowers 

 on the same tree I am unable to say, and only an elaborate 

 inquiry could decide. It is, however, obvious that, owing to the 

 prolonged period of flowering of single trees, there are abundant 

 opportunities of pollination with the pollen of other flowers on 

 the same tree. 



Let us now consider what may be the agents that effect 

 pollination. There is nothing to suggest that the wind blows 

 the pollen from flower to flower or from tree to tree, for all the 

 structural features that characterize wind-fertilised flowers are 

 absent. Though the secretion of nectar is both abundant and 

 long-continued, flying insects do not frequent the flowers ; and 

 indeed the store of nectar is so carefully protected by the natural 

 plug of matted hairs obstructing the corolla-tube that insects 

 could reach it only by biting through the base of the corolla - 

 tube, and this I have never known to occur. There is no doubt 

 that pollination is effected exclusively by small birds. These 

 constantly visit the flowers, hang on the rigid leaf-stalks or flower- 

 stalks, and insert their bills into the corolla-tube to suck the 

 nectar. In sucking the sweet juice the tui may be seen grasping 

 a flower in one foot and turning it round into a more convenient 

 position. In passing from flower to flower the birds cannot 

 avoid bringing pollen from young flowers to older ones, and so 

 effecting pollination. That the arrangement answers its pur- 

 pose is shown by the fairly abundant fruit which the puriri bears 

 even in the neighbourhood of cities, where native birds are now 

 scarce. 



The mechanism for securing pollination is much more com- 

 plete in the puriri than in Rhabdothamnus solandri, the store of 

 nectar is much more copious, and is secreted for a longer time, 

 while the provision for preventing insects from plundering it is 

 most complete. 



Altogether the arrangements herein described constitute one 

 of the most interesting and remarkable adaptations of floral 

 structure to the habits of honey-sucking birds that have so far 

 been detected in our flora. 



