Botany and Zoology. 293 
to suck out nectar from the spur, inserted, as it must be, obli iquely fr 
above, cannot keep the median line at the entrance, but. will take.the the 
right or the left of the protuberance, as may happen, and so will slide 
into the disk-bearing groove of that side, The structure of the disk- 
bearing portion of the column answers, perhaps, to what is expressed by 
indley’s vague character af Gymnadenia, “ rostello complicato,” and is 
tanthera. But nearly every species has its peculiar iewed 
from the front (on removing the Jabellum), each disk is <r to line an 
oblong cavity or deep groove: — vertically from above, this appears 
asa ring with the front edge cut away, or as something more than a 
semicircle, lined by the thin besa disk. On ale a delicate bristle . 
a from above into the spur, taking either si the protuber- 
ce of the labellum, the bristle will ster enter the iscal groove 
buns above, as a thread enters the eye o le, or, if presented more 
obliquely from the front, will slide into the pe when, as it enters the 
spur, it is raised, as it must be, to a more vertical position. The disk clasps 
the bristle, adhering by its sticky surface, and is withdrawn with it along 
with the attached “poilinium, No good observations were made as to 
any movement of the stalk of the proves on the disk when thus ex- 
tricated, nor as to its application to the sti 
It is evident that in this species self- fertilization cannot occur, that only 
one pollen-mass will be likely to be extracted at one visit of an insect, 
and that ~ will doubtless be conveyed to another flower to impregnate 
its stigm 
punters tridentata, Lind|._—Examinations of flower-buds and open 
flowers, July 27-30, substantially confirm those of the previous year, 
ig are recorded in this Journal, vol. xxxiv, p. 426, and in a foot-note 
60. The flowers before expansion are horizontal and somewhat 
shading, so that the poi of pollen, which spontaneously detach theme 
selves from the pollen-mass, may fall out when the anther-cells open. 
The anthers dehisce before the flower-bud is full grown, or at least four 
or five days before the flower opens. In every instance when the flower 
has naturally opened, the anther-cells will be found widely gaping, and 
Several or Bick pollen-packets will be found upon the three “ stigmatic 
process nto which their pollen-tubes will have copiously and deeply 
penetrated. "This, ree is the case two or three days before the blos- 
Som would have opene 
three processes pale so remarkably agen in this species, and 
they so strikingly miso ds 250 Saneernnll, an rance morpholog- 
ically, three elongated or te stigmas, that the eos would not be re- 
r as a congener edly the type of Brown’s genus Gym- 
nadenia, G. flava etal ‘and G. nivea Lindl., Pastis I have examined 
only in dried specimens, however, present interm n 
flava, there are Au strong sti so a Breer laterally beyond the 
by the side of the base of eac middle or rostellar 
lobe is hidden by the —— detheroulla and is functionless. In 
G. nivea, the rostellar lobe is minute and hidden ; the conspicuous lateral 
lobes are linear and deflected forwards so “ 4 lie along the border 
the base of the labellum on each side. Ido not know whether they 
