166 THE JOURNAL OF BOTANY 



minutes another worker comes along, and leaves short-styled pollen 

 on the long-styled pistils, not hitherto cross-fertilized though 

 perhaps visited, by beginning its work with another flower and 

 visiting them in another order. 



Until March 11th I could not fully make out how the pollen 

 from long-styled was transferred regularly to short-styled flowers. 

 After watching some hours about midday I discovered it was 

 brought about in the following manner, thanks to the breeze 

 blowing at the time : the lashing of the flowers on their stalks in 

 the spring wind and the whipping caused by the visits of the 

 bees shake the ripe pollen from the stamens on to the under side 

 of the long-styled pistil and to the corolla-mouth. In rarer cases 

 it appears even to reach the petal disc, for I have observed bees 

 collecting it there. Two days later the rough westerly sleet- 

 showers filled the petal discs of the long-styled upright flowers in 

 my garden with water. In almost every case the pollen of the 

 long-styled form was floating on the water. Yet I have every 

 reason to believe that the plants in my garden have remained un- 

 visited. It is quite unusual to find long-styled pollen on the discs 

 of long-styled plants in the Manor House garden when the bees 

 are working. From the shape of the stigma and its place in 

 regard to the opening of the corolla in the long-styled form, and 

 the closely felted hairs of these parts, only in rarely exceptional 

 cases have a few globules of water entered into the flowers, but 

 the pollen escapes all the same. 



When a bee's hgula is inserted into a long-styled flower a 

 certain proportion of pollen is withdrawn. It may come from 

 the lower side of the stigma or the corolla-mouth side, or it may 

 be gathered from the petal disc, if it has been thrown out by the 

 whipping of the flowers. The bee's next visit may be to a short- 

 styled form. Its first action is to force its ligula amongst the 

 stamens, and by doing this a certain proportion of the long-styled 

 pollen reaches the sliort-styled pistil buried in the corolla. The 

 bee then withdraws its ligula, and begins to collect the pollen on 

 its legs. No doubt the thrusting in of the ligula when it first 

 arrives greatly aids in the dispersal of the pollen in the short- 

 styled form. A certain proportion of the long-styled pollen 

 even reaches the short-styled stigma, for so far as I can learn, 

 unless bees are very plentiful, the crossing from the short to the 

 long is much more regular and bountiful than from the long to 

 the short-styled form. As each flower in the. early days of the 

 season is carefully and systematically worked at in most places 

 by hive-bees, I have no doubt both forms are fully cross-fertilized. 

 Later on, when hive-bees have a wider sweep of floral species, the 

 Primrose is more or less neglected by them. This explains, no 

 doubt, an observation I made as far back as 1877, that the seed 

 from the early wild Primroses is more fertile than from the later 

 flowers. 



While the bees were working the Primroses this season, other 

 hive-bees were visiting the other plants named above, with the 

 exception of Galanthus and Eranthis, which were past flower. In 



