THE STORY OF THE D AVIDIA 291 



mature and change to brown with age. The flowers 

 and their attendant bracts are pendulous on fairly 

 long stalks, and when stirred by the slightest breeze 

 they resemble huge butterflies or small doves hover- 

 ing amongst the trees. The bracts are somewhat 

 boat-shaped and flimsy in texture, and the leaves hide 

 them considerably, but so freely are they borne that 

 the tree, from a distance, looks as if flecked with 

 snow. The bracts are most conspicuous on dull days 

 and in the early morning. 



Later, I found two other trees in the same neigh- 

 borhood and, in localities varying from fifty to one 

 hundred miles apart, eight others. These eleven 

 trees were carefully watched through this anxious 

 year of the Boxer trouble; they fruited freely, and in 

 November I garnered a rich harvest of seeds which 

 were despatched to England where they safely ar- 

 rived in due course. In 1901, when on an expedition 

 through the northwest of Hupeh, I discovered the 

 Davidia in quantity and more than a hundred trees 

 became known to me. From these hundred trees I 

 did not secure a hundred seeds, and during subse- 

 quent visits to China extending over a decade I never 

 again saw Davidia fruiting in the manner it did in 

 1900. The fruit may be likened to that of a walnut, 

 but is more or less ellipsoid, or, more rarely, roundish 



