854 THE POLLINATION OF GOODENIA CYCLOPTERA. 



and does not become viscid till it has exceeded the edge of the 

 indusium. All this time, prior to its viscidity, it has been press- 

 ing out the ripened pollen, and has become quite free from this 

 load before it becomes viscid. Another insect visits the flower, 

 and, probably, brings along some pollen from a previously visited 

 flower: and, on entering the same central passage, gets the pollen 

 brushed off its back by the ciliated indusium and on to the viscid 

 surface of the stigma. 



From these observations, it is hardly possible to draw any 

 other conclusion than that the flowers of G. cycloptera are cross- 

 fertilised. To satisfy myself further on this point by practical 

 means, I tied some gossamer-bags over several bunches of flowers 

 in the open bush-lands, ^vhere they would have every chance 

 from natural causes, and vvatched for results; in every case, the 

 ovaries withered, and fell off the peduncles with the flowers. In 

 no case was there a single ovary or ovule developed; while the 

 ovaries of the neighbouring flowers were developed and bore 

 seeds. From the fact that these flowers were enclosed, not singl}^, 

 but in bunches of three or four, in the bags, it would appear that 

 not only are the flowers of this species cross-fertilised, but, also, 

 that there is ground for very strongly suspecting that the pollen 

 of one flower does not fertilise the ovules of another flower of the 

 same plant; and this seems the more reasonable on account of 

 the flowers being protandrous. This question, however, I leave 

 till further opportunities of observation offer. 



EXPLANATION OF PLATE XCIIL 

 Fig. 1. — Upper part of corolla, and two stamens removed, to show method 



of collecting pollen. 

 Fig. 2. — Same, but after stamens have withered, and the indusium has 



become filled with pollen. 

 Fig. 3. — Bud shortly before opening, to show bending of .style and corolla 



apart, and the former becoming exserted. 

 Fig. 4. — Same, with portions removed to show the indusium pressing 



against the auricle of the remaining lobe. One auricle removed. 

 Fig. 5. — Flower with sides removed, to show the action of an insect iu 



forcing itself under the indusium. (This was caught in situ). 



(Figs. 1-4 much enlarged.) 



