is simply another of the many perplexing coincidences 
of botanical exploration. 
Notwithstanding the intensive collecting that has 
been done in Costa Rica, Stelis glossula, so tar as my 
records indicate, escaped detection following its publica- 
tion in 1870 until 1984, when it was rediscovered by 
Lankester. It isindeed noteworthy that both Stelts glos- 
sula and S. inaequalis, which constitute a distinct sub- 
section of the Middle American stelids, should exhibit a 
high degree of localization in their distribution. 
To account for the rarity of Stelis glossula and its 
recent rediscovery in Costa Rica, following years of in- 
tensive botanical collecting there, it is only necessary to 
consider the limited distribution of many orchids and the 
chances against rediscovery if the restricted areas in which 
they grow are not visited by collectors. It would seem 
reasonable to suppose that Lankester collected his speci- 
mens of S. g/ossula in or very near the type locality and 
the species may continue to be missing from collections 
until that locality is again explored or another locality 
where it grows is discovered. Always, however, in dis- 
cussing the rarity of certain epiphytic orchids, it should 
be remembered that they usually establish themselves on 
the limbs of lofty trees in dense tropical forests, a pecu- 
liarity in their occurrence which limits the frequency with 
which they are detected and results in rarity being as- 
cribed to them although in reality they may be common. 
Stelis glossula and S. inaequalis differ fundamentally 
from the species which Lindley grouped together in his 
section Labiatae, in that the secondary stems are almost 
obsolete and the plants in their habit resemble species of 
Masdevallia. Typical species of the Labiatae have elon- 
gated stems with the secondary stems well developed and 
usually widely separated. Furthermore, the labellum in 
flowers of typical examples of the Labiatae differs mark- 
[150 | 
