294 Trmuactions. — Botany. 



fertilization through the agency of insects ; and there can be no doubt that 

 if the flowers were regularly visited by suitable species this would inevitably 

 take place. But, from some reason — probably from the want of sufficient 

 attraction — insects seldom visit the flowers. For the last seven years, I 

 have made it a practice to watch beds of this Orchid, and save on two occa- 

 sions I have never seen winged insects enter the flowers ; and in both these 

 cases the poUinia were not removed. It should, however, be mentioned that 

 a minute thrip-like insect is sometimes abundant on the pollen, on which it 

 probably feeds ; but it is much too small to be of any service in removing 

 the pollen from flower to flower, although it may be useful in another way. 

 But although insects have not been actually observed in the act of removing 

 the pollen, I have yet been able to collect evidence proving that they occa- 

 sionally, though very rarely, do this. Thus, in November, 1876, seventy- 

 five flowers were examined, and two had lost their pollinia — evidently 

 removed by some insect. In November, 1878, 103 flowers were examined, 

 and two of these had their pollinia removed. Lastly, in November, 1879, 

 out of forty flowers three had the pollen-masses missing. In three instances 

 I have also found pollen scattered over the stigma, the pollinia being still 

 intact in their places at the back of the rostellum. These facts are quite 

 sufficient to show that cross-fertilization does occasionally, though very 

 rarely, take place. 



Few of our indigenous species mature seed so abundantly as Thelymitra 

 longifolia, almost every flower producing a ripe capsule. Taking this in 

 connection with the facts mentioned above, it is obvious that we have to do 

 with a case of self-fertilization. The mode in which this is effected appears 

 to be as follows : — The upper part of the stigma is thin and membranous, 

 and has its margin slightly revolute, even when in the bud. After expan- 

 sion this rolling back is carried to a greater extent, so that the edge of the 

 stigma, and even a small portion of its anterior surface, comes into contact 

 with the pollen-masses hanging directly behind it. Pollen-tubes are at once 

 emitted into the substance of the stigma, usually so rapidly that before a 

 flower has been expanded more than a single day the pollinia are glued so 

 firmly to the margin of the stigma that they could not be removed by insects, 

 even if they visited the flowers. Pollen-tubes are also frequently emitted 

 into the upper part of the posterior surface of the stigma. Besides this, it 

 often happens that the tops of the pollen-masses (which, as we have seen, 

 slightly overtop the stigma) are broken down by some means, either by the 

 shaking of the flowers by wind or by the minute thrip-like insect already 

 mentioned, and the pollen scattered over the front of the stigma. By one 

 of these methods, or by both combined, I believe that the flowers are regu- 

 larly self-fertilized, and the perpetuation of the species secured. 



