SELF-FERTILIZATIOK IK TROPICAL ORCHIDS. 547 



observation is of a species unknown to me, and nearly related, if 

 not belonging, to the genus Glirysoglossum. It carries the contri- 

 vances for self -fertilization to the utmost limit. It is terrestrial ; 

 has long pointed fleshy leaves, and produces a spike about one 

 foot long, with alternate flowers at short intervals. It is not 

 only self-fertilized, but cleistogamous. I first found it at a 

 height of about 3000 feet, where I observed that all the flowers 

 were fertilized, but to all appearance none had ever opened. 

 I brought several plants home and planted them in pots, where 

 they at length threw out new spikes, which appeared to be 

 perfectly healthy. The flowers remained as buds nearly two 

 weeks, during which the ovary daily increased in size ; then they 

 shrivelled and dried up, while the ovary enlarged to its full size. 

 On opening it, I found it to be well filled with seeds. The labellum 

 is beautifully marked with lines of purple- carmine ; the column 

 with carmine and yellow ; but no insect could ever be fascinated 

 and allured by its painted whorls. 



The last example is a species of Goody era, G. grocer a, Lindl. 

 I have found a very large number of the little flowers to be self- 

 fertilized; the pollinia from some cause emit the pollen-tubes while 

 still on the anther, which, travelling by way of the edges of the slit 

 of the rostellumj penetrate the stigma and fertilize the ovary. The 

 fiWers w r ere too small for thorough examination with the meaus at 

 my disposal. The labellum is produced into a kind of nectary. I 

 could not detect whether it contained nectar or not, but its walls 

 are studded with numerous long, narrow, knobbed glands. 



Besides these, Cryptostylis Arachnites and several other species, 

 which I have not yet been able to identify, presented similar or 

 allied modifications for securing self-fertilization. 



From these observations I am inclined to think that a great 

 number of Orchids are self-fertilized. The great family of the 

 Vandese, however, seem rarely, if ever, to be self -fertilized ; they 

 are either cross-fertilized, or altogether fail to set seed-capsules. 

 Terrestrial Orchids seem more liable to self-fertilization than 

 epiphytic species. 



Summary. 

 In the foregoing pages I have drawn attention to the fact :— (1) 

 that a number of Orchids with showy flowers never set any seed- 

 capsules ; and (2) have given additional examples to those men- 

 tioned by Mr. Darwin of Orchids of different genera being invari- 



