stant) which usually formed the major key to separation, 
and investigation shows these characters to be wholly 
unreliable. For instance, the size of the plant, the width 
of the leaves, the relation of the height of the inflores- 
cence to that of the leaves, and the racemose or panicu- 
late form of the inflorescence were erroneously considered 
to be of great weight. Contributary marks of distinction 
were taken in the size and termination of the floral seg- 
ments and in the exact form of the petals and lip. 
I have come to the conclusion, concurred in by that 
astute student of orchidology, Mr. Leslie Garay of To- 
ronto, Canada, that Polystachya takes its place among 
such strikingly polymorphic groups as Epidendrum, 
Spiranthes and Habenaria. As a consequence, it has 
seemed advisable in the interests of truth and simplicity, 
to make the above reductions, now that an abundance 
of material is available. 
Mr. Garay points out that his ‘‘microfilm copy of Poep- 
pig’s Nova Genera bears a pencil mark of unknown ori- 
gin which indicates that on plate 113 the habit sketch B 
(Encyclia polystachya) and the details A belong together 
[contrary to the citation], also the habit sketch A and 
the details B.’’ Indeed this interpretation proves to be 
true, judging from Cogniaux’ description of Polystachya 
nana in Martius Flora Brasiliensis, Vol. 8, pt. 4 (1896) 
319, after his having seen the Poeppig type of //ncycla 
nana. However, his description of the lip of Hnecyclia 
nana agrees well with the drawing of the lip of Hncyclha 
polystachya. 
After making allowance for the well-authenticated fact 
that the mid-lobe of the lip of this group of Polystachya 
is often slightly retuse at the broadly rounded apex, it 
appears certain that the concepts Stelis foliosa Lindl., 
Polystachya cerea Lindl., Polystachya clavata Lind|., and 
Polystachya caracasana Reichb.f. (all represented in the 
[ 53 | 
