366 THE ORCHID REVIEW. 
with the median petal, or lip, of which they form the two side lobes, and as 
A I forms part of the column and the lateral petals have no such append- 
ages, it is clear that A 2 and A 3 have wandered from their proper course. 
The filaments also are more confluent with the column, and the anthers 
more reduced; while the pollen has become viscid, and one lobe of the 
stigma (s 3) suppressed. In short the flower has become very irregular and 
highly specialised. In Monandre further changes take place. A I is now 
invariably the fertile anther, while a 1 and a 2 are reduced to staminodial 
appendages, as wings or teeth of the column, or they may be entirely 
suppressed. The rostellum also appears for the first time, and, corellated 
with this, the pollen grains now become variously aggregated together. 
The effect of these and other progressive modifications has been to render 
them increasingly dependent on particular insects for the fertilisation 
of the flowers, 
The foregoing remarks not only explain the structure of an Orchid 
flower, but also help us to understand many of the malformations which 
from time to time appear. Take for example a four-flowered spike of 
Cattleya labiata which has just appeared with Messrs. Hugh Low & Co., 
every flower having three perfect anthers, but the lips developed as ordinary 
petals. This is a case of reversion. The two stamens A 2 and A 3, instead 
of uniting with the lip and being modified into petaloid staminodes, have 
being carried up with the column and developed into perfect anthers, 
precisely as in the case of A 1, the column now being perfectly straight, and 
the stigmas abortive. The stamens having become dissociated from the 
median petal that organ resembles the lateral petals, and thus the flower 
becomes perfectly regular. A similar flower has previously been 
recorded (Vol. II., p. 358), but in that case a second flower on the 
same raceme had two of the petals transformed into lips. The column 
was imperfect, and without either anther or rostellum, the latter 
having reverted to its original state, and being confluent with the two other 
stigmatic lobes into a terminal stigma. In this flower it is probable that 
A ti had wandered from its proper course, and becoming confluent with a 
lateral petal transformed it into a lip. These two cases are very instructive. 
In the case of Cypripedium insigne with three lips, a 1 and a 2, instead 
of going up with the column and forming perfect anthers, had become con- 
fluent with the two lateral petals, and were transformed into petaloid 
staminodes, as in the case of the normal lip. This, however, cannot be 
considered as a case of reversion. The flower of the same species, with the 
side lobes of the lip alone developed, owed its peculiarity to the suppression 
of the median petal. 
The wings 
of the column of Monandre we have already seen to be 
staminodes, 
€s, representing ar and az, but these occasionally bear perfect 
anthers, in which case the three anthers are situated on the back of the 
