METAMORPHOSIS. 



145 



ing the accepted theory and to support that advanced 

 by Oriiger, are the following : 



1. There is no more a priori necessity for regarding 

 the labellum of orchids as a compound structure than 

 there is for holding the same view with regard to a 

 perianth-leaf of complex organization in any other 

 plant, e. g. the outer perianth -leaves of Iris, the 

 nectary-petals of Aquilegia, or the posterior sepal in 

 Delphinium, or Aconitum. No special hypothesis is 

 required to explain the presence and structure of the 

 labellum. 



2. Wydler found in the bird's-nest orchid (Neottia) 

 that the stamens A 3 and A 3 were present, and incom- 

 pletely fused with the column, yet " the perianth was 

 quite unchanged and retained its irregularity." This 

 fact shows that the labellum can be produced without 

 the intervention of stamens A 3 and A 3 . 



3. Hansen described in the 6 Gardeners' Chronicle,' 

 and Masters and Penzig describe the same thing, a 

 flower of Phajus g randif alius and one of Odontoglossum 

 crispum possessing three labella ; the stamens A 2 and 

 A 3 not only being present and entirely distinct from 

 the labellum, but being themselves transformed into 

 labella. 



The flowers of two varieties of Dendrobium nobile 

 have been described above in which three labella were 

 present, the " column" being normal. 



Cases like these are sufficient to entirely dispose 

 of the hypothesis of Brown and Darwin ; the facts 

 indeed speak for themselves. It might be supposed, 

 possibly, by some enthusiasts for the orthodox view, 

 that the normal labellum, in the absence of the help of 

 its usual lieutenants A 2 and A 3 , had obtained, for its 

 development, that of a? of the inner whorl, while the 

 two extra labella had similarly arisen through the 

 timely resurrection of a } and a 2 . But apart from the 

 fact that a labellum, on the accepted theory, requires 

 two stamens for its adequate formation, this forlorn- 

 hope-position is at once destroyed by the facts observed 



vol. n. 10 



