MB. G. BENTHAM ON OECHIDE^I. 289 



cohaerens. Labellum basi columnae adimtum, ultra antheram varie pro- 

 ductum v. appendiculatum. 



Tribus 5. Cypripedie^e. Antherae 2, ad latera rostelli v. styli sessiles v. 

 stipitatae, polline granuloso ; anthera postica in antheridium polymorphum 

 mutata, rarius perfecta v. omnino deficiens. 



I now proceed to enter into a few explanatory details, takiug 

 the several tribes in the above order. 



Tribe 1. Epidendre^. 



This tribe is formed of the union of Lindley's Malaxidese and 

 Epidendrese, which, as already observed, he had distinguished by 

 the absence or presence of a caudicle to the pollen-masses ; but 

 owing to the vagueness of the meaning attached to the term 

 caudicle, and the real uncertainty in many cases as to the substance 

 which often connects the pollen-masses, there are so many genera 

 whose place in the one or the other tribe has been a matter of 

 doubt, that Lindley himself had suggested the consolidation of 

 the two, and their subdivision on other principles. This process 

 he unfortunately never carried out in detail, although he gave some 

 indication of it in his lists of genera in his ' Vegetable Kingdom.' 



As a whole, Epidendrese are chiefly distinguished from Vandese, 

 the other great tribe of Orchidese with waxy pollen-masses, by the 

 distinctness of the two anther-cells, which are always parallel, or 

 nearly so, and after discharging their pollen leave their margins 

 or valves prominent within the anther-case, and by the removal 

 of the pollen without carrying off any scale-gland or stipes formed 

 by a layer or plate detached from the rostellum. This character 

 is, in the great majority of genera, well marked and readily ascer- 

 tained ; but in some instances it requires very careful observation 

 not to mistake it, and sometimes may really be rather uncertain. In 

 coming to the following conclusions, I have been guided in the 

 first instance by Darwin's clear exposition of the results of his 

 careful study of the process of fertilization in a few leading 

 genera ; and I have followed them up by the observation of such 

 species as I have been able to procure in a living state, and by 

 the close examination of buds and open flowers in dried specimens 

 of a great majority of the genera, and generally of many species 

 of the larger genera. I am fully aware, however, that in this 

 respect dried specimens often give but very unsatisfactory data. 

 In those gathered wild the pollen is often already carried off by 

 insects from flowers but just expanded; and it is so readily dis- 



