G. M. Thomson. — On the Fertilization of Flowering Plants. 285 



and the meclianism by which cross-fertihzation is effected is the most 

 perfect of its kind. The flowers are produced in few-flowered racemes 

 seldom exceeding an inch in length, and more or less hidden by the leaves. 

 They are much more regular than the majority of Orchid flowers, are 

 greenish in colour-, with a few purple lines on the labellum, and do not ex- 

 ceed one-tenth of an inch in diameter. 



At first sight I thought that owing to their inconspicuous appearance 

 they must certainly be self-fertilized, but besides being slightly fragrant, I 

 noticed that they secreted a considerable amount of honey between the base 

 of the column and the fleshy, ridged labellum, and this caused me to look 

 more narrowly into their structure. The four poUinia are united into two 

 almost globular masses, which are attached by a caudicle to a broad flat disc 

 fixed to the rostellum. If this be removed fi'om the anther, which is at the top 

 of the column, it at once commences to contract, and thus causes the pollinia 

 to be depressed to a nearly horizontal position. This depression is almost 

 identical with that which occurs in the British Orchis mascula, as described 

 by Darwin,* but there is a somewhat different action in our species, in that 

 the two masses of pollinia separate slightly at the same time. The time 

 taken by this contraction and depression of the caudicles, was about ten 

 seconds. If these were attached to the proboscis of a small insect, they 

 would on their first withdrawal from the anther be in such a position as to 

 strike the rostellum of the next flower they visited, but this is obviated by 

 the depression of the caudicle, so that, in the short interval of time men- 

 tioned, they are so placed as to project into the deep and somewhat two- 

 lobed stigmatic cavity under the rostellum. 



In my former paperf I stated that Chiloglottis cornuta seemed exclusively 

 adapted for self-fertilization. This I am now enabled to verify, for a 

 number cultivated indoors, and covered by a hand-glass during the flower- 

 ing season, produced a fine full capsule from each flower. 



Of the Thelymitras, T. longifolia is very frequently self-fertile, but promi- 

 nent forms are no doubt crossed by insects. T. unijiora, on the contrary, 

 which produces very brilliant blue flowers, is chiefly dependent on insect 

 aid, and is a great honey-producing species. 



Nat. Ord. lEiDEiE. 



This order is only represented in New Zealand by the genus Libertia, 

 two species of which are found in Otago. 



L. ixioides is an extremely common plant, particularly in dry, sandy 

 situations. The white, hermaphrodite flowers have no scent, but in bright 

 sunny weather secrete a considerable amount of honey, and are very con- 

 spicuous. I have not detected any difference in the time of ripening of the 



* "Fertilization of Orchids," p. 6. f " Trans. N.Z. Inst.," Vol. XI., p. 424. 



