BOTANICAL SUBJECTS. 79 



a segment of the inner whorl, and not of the outer whorl, as in 

 Disa, 



Therefore by transposition of the petals of Angrcecum into 

 sepals, we may obtain a Disa. But the reader might say, is it 

 conceivable that the two petals of Disa which are enclosed in its 

 scooped sepal can be made to become exterior ? Quite conceivable. 

 The reader should, however, note that I said an Angrsecum might 

 easily become a Disa, and not D. yrandijlora ; for if he will take 

 the trouble to look up Disa Macrantha, he will find it has the 

 following structure : — three sepals, the upper one of which is large, 

 like a cowl with a long spur, and that the two upper petals are 

 not ensconced in the cowl. It would therefore seem that when 

 the Jiower is in hud^ and the segments in the stage of nipples, 

 nothing would be easier than a transposition of petals into sepals, 

 in grooving, and the modification of an Angrsecum into a Disa. 



I am aware that in Disa spathidata, opposed to the spurred 

 cowl, there is a much modified petal, which may represent the 

 labellum of other orchids, but it is the spurred segment of these 

 orchids which appears to me to be the morphological homologue in 

 both. In the one case it is called a petal, in the other it is called 

 a sepal. 



I have elsewhere* shown that a spur can be formed all of a 

 sudden, without being inherited, but when once formed we cannot 

 avoid taking into consideration the fact that it is more readily 

 inherited than newly formed. 



Here, then, I think, we have two genera which are almost 

 interchangeable by the mere transposition (when in bud), of inner 

 into outer segments, or vice versa, and I should not be at aU 

 surprised to hear that Disa and Angnecum would inter-breed, 

 more especially as the Messrs. Veitch have been so successful in 

 breaking down anatomical genera into physiological varieties, 



I have in store, however, better proof than all I have adduced in 

 support of my statement, viz., that transposition of petals to sepals 

 is not only very probable, but that it does actually occur. 



The hybrids of Gladiolus Gandavensis have exercised my 

 thinking faculties for some time to account for a phenomenon that 

 so frequently occurs among them. It is this : The Gladiolus is 

 an irregular Irid, with three segments above and three below. 



See chapter on " Spurs." 



