852 THE POLLINATION OF GOODEMA CVCLOPTEHA, 



to protrude the indusium through tlie slit between the two upper 

 lobes of the corolla"; and, then remarking that the indusium 

 re-enters the corolla, came to the conclusion that this species is 

 self-fertilised. But he records the remarkable feature, that "the 

 same set of organs which, in G. ovata, prevent self-fertilisation, 

 in G. hederaced ensure it." 



Now, if this divergence in the mode of pollination and fertili- 

 salion in these two species has been rightly decided by these two 

 careful observers, one would be inclined to conclude, that G. 

 ovata, being an erect, and G. hederacea a decumbent species, then 

 other erect species of this genus would be likely to be cross- 

 fertilised, and the decumbent species self-fertilised. I do not, 

 however, find this to be the case in G. cycloptera, the decumbent 

 species under consideration. In fact, in this species I find diver- 

 f^ences from both the above species, as observed and noted by 

 the authors mentioned ; principally, in the fact, that the}^ 

 state the indusium, in both the species noted, exserts itself from 

 the flower, and then, becoming inflected, re-enters it. In the 

 case of G. cycloptera, I find that the style bends upon its ventral 

 side, and thus protrudes its middle portion through the upper 

 division of the corolla; but that the indusium, at the same time, 

 and by reason of this exsertion, and aided also by the tendency 

 of the corolla to bend back, presses hard against the auricles, 

 which, by reason of their excessive induplication, are constitu- 

 tionally enclosed well within the bases of the lobes of the upper 

 part of the corolla. Hence, the indusium of G. cycloptera does 

 not at any time protrude itself from the flower, as is stated to be 

 the case in the species compared above. 



Now, in inquiring further into the pollination of G. cyclo)>tera, 

 I think it will be found, contrary to the expectations expressed 

 above, that, though it be a decumbent species, it is, after all, 

 cross-fertilised. The method of pollination in this species may 

 be thus stated. When the bud is but half-grown, the stamens 

 are connivent around the style, with the base of the anthers just 

 pressing against the indusium; the style at this stage develops 



rapidly from observations most carefully made, I ha\ e calculated 



the rate of growth to be one line an hour — such a rapid move- 



