158 ON THE FERTILIZATION OF GOODENIA HEDERACEA, 



except that the ledge representing the fnture indusium has grown 

 up higher in the centre and extended to the ends (see fig. 2}. In 

 a still farther advanced bud the only change in the relative position 

 of the style anthf^rs is that the latter are slightly higher than 

 before, but in the style itself, a great deal of alteration is percep- 

 tible, the indusium being now higher than the stigma, and the 

 edges showing the hairs which are so striking a feature of the 

 mature organ (see fig. 3). We now examine a bud nearly ready 

 to open. Here we find the anthers full grown but not mature, 

 and leaning over the indusium (fig. 4) their ba&es being on a level 

 with the top of that organ, and the upper part of the filament 

 developed into a point which projects beyond the anther itself (see 

 fig. 5). The indusium is develoi)ed into a deep cup, the edges 

 being densely clothed with short thick hairs, and the outside 

 inferior surface having a quantity of longer and thinner hairs on 

 the centre (see fig. 6). At the bottom of the cup is the stigma, 

 now almost ready to receive the pollen. For the next stage, it is 

 necessary to choose a bud just beginning to show slits at the sides. 

 On removing the corolla the anthers will be found clasping the 

 indusium and the points (figs. 7 and 8) turned over into the cup 

 of the indusium, which is quite full of pollen. The anthers will 

 be seen to be quite empty, and if the parts be exposed to the air 

 for a short time (as would happen naturally by the fuller opening 

 of the bud) the filaments contract and twist and the anthers shrivel. 

 It will immediately strike the observer that the style must have 

 lengthened considerably and rapidly, as in all younger specimens 

 the bases of the anthers were on a level with, or above the cup, 

 and here it is the points of the anthers which are level with it. 

 Another noticeable feature is the packing of the pollen into the 

 cup — a point of which I shall have more to f-ay presently. Our 

 next step is to examine a fully open flower. The basal poi-tion of 

 the style is bent upwards so as to protrude the indusium through 

 the slit between the two upper lobes of the corolla. The anthers 

 are entirely empty, withered, and bent back through the slit, so 

 that they are outside of the flower — with them we have done. The 

 slit between the up[)er divisions of the corolla widens in the centre, 



