258 ON SOME ORCHIDS OF THE SOUTH OF FRANCE, 
the movements were effected in about 3 minutes. The pollinia of 
sueceeding flowers became gradually more and more efficient, and 
fertilized a fresh spike in the ordinary way. This shows that the 
pollinia may be removeable and useful after the spike on which 
they are is incapable of fertilization. 
Fig. 2 represents a spike of Serapias cordigera (Linn.), a 
plant which has two pollinia united on one gland, as in the pre- 
ceding, but possesses a complex and most interesting fertilization. 
On their withdrawal the masses are bent back away from the 
stigma (b), but quickly reverse and accomplish their depression 
and contraction as in Orchis hircina. The stigmatic cavity is ex- 
ceedingly narrow and obseure; so that, though the column is en- 
closed in a somewhat tubular chamber formed by the upright an- 
terior lobes of the labellum and the hood-shaped coherent seg- 
ments of the flower, the pollen-masses on an insect's head would 
be very apt to miss coming in contact with its viscid surface at all. 
To remedy this, the guiding-plates (a, fig. 2) are raised into a 
two-walled glabrous trough, in whieh the masses slide without 
fail against the stigma. I secured a specimen of an insect ( Cera- 
tina albilabris) in the act of touching the stigma of this plant with 
the yellow pollen of Serapias lingua, two pairs of pollinia of which 
were fastened on its head ; the gland of one pair partially covered 
the right eye. ' 
In the case of Xylocopa violacea and Ceratina albilabris, therefore, 
we find the glands attached to their heads, not their proboscides ; 
and it seems probable, as the height at which the pouch stands is 
most accurately in relation to that of the insect fertilizer, and 
as the flat glands are not easily removed by any narrow object or 
slight degree of pressure, that the elevation of the pouch is a fea- 
ture of very great importance to the plants. My last observation 
concerns Ophrys Scolopax (Cavanilles), a plant very analogous to 
O. Arachnites. This plant appears under two forms in the two 
localities where I have obtained specimens. At Mentone I never 
saw any tendency to self-fertilization, but all the spikes of a large 
bundle sent me at Cannes were so without exception. This ma- 
terial difference between the two is accomplished by a very slight 
bend in the anther-cells, which are prolonged into a beak of va- 
riable length, in the case of the self-fertilizing blossoms. It is a 
remarkable coincidence, that at Mentone the Bee Ophrys is 
scarce, and at Cannes very abundant. So, within thirty miles of 
one another, we have one spot where self-fertilization is in full 
action, and another where it is, as far as I am aware, unknown. 
