DAVIDIA 477 



The winter of 1908-9, which was the severest since its introduction, and 

 injured many reputedly hardy plants, did not affect the Davidia to the 

 extent of a single bud. It appears to thrive best in a loamy soil, and 

 loves abundant moisture. It can be propagated by cuttings made of 

 short side twigs when the wood is becoming firm. 



As a garden tree the Davidia is one of great promise. It stands 

 quite apart in its distinct beauty from all other trees that can be 

 grown in this country. Mr Wilson has many times told me of the 

 wonderful aspect of the tree as seen in its native wilds, laden from top to 

 bottom with enormous white bracts. It flowered at Coombe Wood in 

 May 1911, and from a spray of the tree the accompanying figure \^as 

 made. The leaves were farther advanced than is usual with the tree 

 when in bloom at home. Davidia stands much apart in the vegetable 

 kingdom ; its nearest ally is considered to be Nyssa. 



There remains to be considered the question as to whether there be 

 one or more species of Davidia. The original tree discovered by David 

 had leaves as described above, viz., white-felted beneath. But the trees 

 introduced by Farges, and later by Wilson, have leaves almost or quite 

 smooth, and they differ from each other in one having leaves glaucous 

 beneath, the other yellowish green beneath. On the strength of these 

 differences Mr L. A. Dode of Paris has founded two additional species : 



1. D. involucrata, Baillon. Leaves whitish, felted beneath (the original type discovered 



by David). 



2. D. Vilmoriniana, Dode. Leaves smooth and glaucous beneath (this is the single tree 



raised by Mr de Vilmorin from Farges' seed). 



3. D. laeta, Dode. Leaves smooth and yellowish green beneath. (To this belongs the 



great number of plants raised by Messrs Veitch from Wilson's first seeds ; as 

 stated above, it flowered at Coombe Wood in May 1911 the stamens white, with 

 black anthers.) 



Whether the distinctions on which Mr Dode's species are founded 

 will prove to be permanent or not remains to be seen. I do not think 

 that the differences between Nos. 2 and 3 are sufficient to constitute a 

 species ; but those between that pair and the original D. involucrata of 

 David's finding are certainly very marked. Mr Wilson believes the 

 hairy- and the smooth-leaved trees to be separate and distinct, and he 

 considers the former to be the more handsome. He tells me he found 

 them growing commingled so they are not geographical varieties. 

 During his second journey for Messrs Veitch in China (1903-5), he sent 

 home seeds from the hairy-leaved tree (the true D. involucrata) gathered 

 by his own hand, but it is a curious fact that the young trees raised from 

 them at Coombe \Vood at first showed very little of the hairy character of 

 their parent, but they are becoming conspicuously hairy now. If they 

 should prove eventually to be distinct, then : 



D. involucrata (true) will have been first introduced by Wilson in 

 1905 ; and 



D. Vilmoriniana by Farges in 1897. D. laeta need not be considered 

 of much importance as apart from Vilmoriniana, but it was introduced by 

 Wilson in 1901. 



