Cheeseman. — On the Fertilization of Thelymitra. 295 



Mr. Fitzgerald, in the introduction to liis magnificent work on Australian 

 Orchids, states that T. longifolia is fertilized in the hud in Australia. This, 

 however, is certainly not the case in New Zealand, save when a long suc- 

 cession of wet weather has prevented the flowers from opening at their 

 proper time. In ordinarily fine seasons I always find that the polhnia are 

 intact and free from the stigma on the expansion of the flower, and come 

 away with the rostellum on its removal. He also states that the flowers 

 open for one hour only in the middle of the day. In New Zealand nearly 

 all the varieties open for a much longer period than this, the chief exception 

 being a blue-flowered form with very long and slender staminodia, ap- 

 parently an intermediate between T. longifolia and T. pidchella. I cannot 

 agree with Mr. Fitzgerald in considering the opening of the flowers " use- 

 less," for even admitting that the stigma had been penetrated by pollen- 

 tubes prior to the expansion of the flower, yet if pollen should be brought 

 from a different plant and placed on the stigma, it would probably have a 

 prepotent influence, and destroy any effect produced by the plant's own 

 pollen. To me it appears that the opening of the flowers is highly im- 

 portant, even if it takes place for only an hour, for it gives a chance of 

 cross-fertilization being eff'ected, and the great value of this is now well 

 established. It is curious that Mr. Fitzgerald's researches should appear 

 to go towards proving that T. longifolia is more exclusively self-fertilized, 

 and the flowers more nearly cleistogene, in Australia than in New Zealand ; 

 for, considering the admitted paucity of insects in New Zealand, and the 

 much damper and cooler chmate, the exact converse of this might have 

 been anticipated. 



The whole case of Thelymitra is most interesting, and at the same time 

 perplexing in the highest degree. On the one hand we find the flowers 

 possessing a viscid rostellum, to which the pollen-masses become spon- 

 taneously attached, and with which they can be removed, — clearly an 

 adaptation for cross-fertilization through insect agency. On the other hand 

 we see that the form and position of the stigma is such that it early comes 

 into contact with the pollen-masses, an end which is also encouraged by the 

 margins of the stigma slightly bending back towards the polhnia ; we also 

 find that in some of the varieties the flowers remain closed for a large part 

 of the day, thus absolutely preventing the access of insects. It is impossible 

 to doubt that these cu'cumstances favour self-fertilization. We are thus 

 driven to the same conclusion that Mr. Darwin has arrived at in the case of 

 the Bee Orchis : — that in the same flower there exist elaborate contrivances 

 for directly opposed objects. To believers in the theory of evolution, the 

 case is not without an explanation. If we could trace back the modifica- 

 tions through which the plant has passed, we should probably arrive at a 



