in Li Hum Mart agon: II. Spermatogenesis. 21 1 
I have often watched hive-bees visiting plants of Lilium 
Martagon in search of honey. The flower when it first opens 
hangs downwards, the petals are quite flat, and the stigma 
is seen below them like the clapper of a bell. It is consider- 
ably longer than the stamens. Later on in the same day the 
petals curl up, leaving the circle of anthers exposed round the 
long pendulous stigma. The three older anthers usually 
dehisce about this time. Next morning all the anthers have 
dehisced, but the stigma still hangs downwards. A bee in 
search of honey climbs in by the petals and clings to the 
stamens while it is sucking. Thus the under side of its body 
gets well dusted with pollen from the versatile introrse anthers, 
but the stigma is untouched. On the third morning, however, 
when the anthers have shed most of their pollen, the style 
bends upwards, and forms a convenient perch on which the 
honey-seeking bee may alight. In this way it is sure to be 
touched with pollen from the lower part of the bee’s body. 
Bees invariably begin with the lowest flower of an inflorescence 
and work upwards. Cross-fertilization is ensured, as in the 
Foxglove, by the proterandry of the flowers and their centri- 
petal arrangement on the stalk. 
The stigma of Lilium Martagon is three-lobed, and each 
lobe is outlined by a depression in the stigmatic surface. It 
is also bisected for a part of its length by a cleft, the three 
clefts meeting in the centre of the stigma and running down 
into the channel of the style (Fig. 29). On either side of each 
cleft is a slight ridge, covered, like the rest of the stigmatic 
surface, with stigmatic hairs. More pollen is commonly 
deposited on the ridges bordering the two upper clefts in the 
upturned stigma than on the ridges of the third. 
It is a laborious business at best to hunt for dividing 
generative nuclei in pollen-tubes creeping among stigmatic 
hairs. I have identified such stages most easily in hand- 
sections from alcohol-material. The stigma is cut from the 
style about a millimetre below the stigmatic surface, and 
the thick slice thus obtained is stained en bloc with a dilute 
solution of Mayer’s haemalum in *i °/ 0 potash alum. It is 
Q 
