286 ME. Q. BEKTHAM ON OECHIDE^I. 



exceed in bulk the pollen-masses themselves. 3. The so-called 

 caudicle in Vandece is very different both in origin and substance ; 

 it forms no part of the pollen, nor even of the anther, but is a 

 production of the upper surface or back of the rostellum, being 

 a prolongation of the so-called gland or detachable disk of the 

 rostellum. Darwin, distinguishing it from the caudicle, proposes 

 to call it a pedicel, which would have been an appropriate term 

 but for its universal use in descriptive botany for the special 

 designation of the pedicel of a flower. The term stipes is equally 

 appropriate, and has not the same inconvenience, for it is 

 generally used as the support of any organ. The presence of 

 this stipes, though general in Vandece, is by no means universal, 

 and traces of it may be found in genera belonging to other tribes. 

 The tribe of Vandece can, however, be maintained with advantage 

 with very nearly the limits assigned to it by Lindley, but founded 

 on other characters ; but the Malaxidece and Epidendrece must 

 be amalgamated, as already suggested by Lindley, and may be 

 divided into several tolerably distinct subtribes. 



Of the three tribes with granular or powdery pollen, that of 

 Oplirydece remains as a very natural and perfectly distinct 

 group as limited by Lindley, but requiring some little modifi- 

 cation of the technical character ; but the separation of Are- 

 thusece from Neottiece has proved to be purely artificial, without 

 even the advantage of a constantly definite distinctive character, 

 although here again, as in Epidendrece, several natural and 

 tolerably well-characterized subtribes may be recognized. 



In the small anomalous tribes Cypripediece and Apostasiece 

 there is no alteration to propose other than their consolidation 

 into a single one. 



The result of a detailed examination of all the genera pro- 

 posed or established of which I could procure specimens, living 

 or dry, checked by published descriptions and illustrations, has 

 been their distribution into five tribes, slightly modified from 

 those of Lindley, and twenty-seven subtribes, of which I now 

 give a short summary of the most essential characters, passing 

 over for the present the exceptional forms. These exceptions 

 will, I think, be found to be very few as connecting the tribes ; 

 but the subtribes are not always so definite, and it may be 

 hoped that a further study of numerous forms of which we 

 have at present only very imperfect materials may lead, in many 

 respects, to considerable improvements in their circumscription. 



