- â78 - 



convenience'', Watt treats this form as a variety. It is frequently found 

 both in India and in China, whereas the former variety (improperly termed 

 „China tea") is scarcely represented in collections from China. 



„Ad 3 (expressly written on niy request). — „lf you cross Manipiir with 

 „stricta you would get a plant that would be difficult to separate from 

 „many of the Ceylon and South Indian teas." 



Indeed, it is far from easy to say anything against this theory; thèse 

 hybridization problems can be solved by way of experiment only, not with 

 the aid of suppositions based on morphological comparison. Nevertheless 

 I venture the following contra-suppositions : 



la. The variety viridis Watt is a composite of heterogeneous constituents, 

 to wit: Indian and Chinese forms. 



\b. The variety lasiocalyx Watt is one form of the Chinese form 

 groiip, being of subordinate systematical importance. 



le. The variety stricta Watt equally belongs to the group of genuine 

 China tea. 



2. The variety bohea Watt is composed of certain races of China 

 tea and actual hybrids (vide 3) between India and China tea. 



3. In stead of: viridis X s^r/c/a = Ceylon hybrid, the hypothesis ought 

 to be read thus: India X China = „hybrid" type 



Ad la. — Apart form the formally contestable fact that Watt applies 

 the name viridis to a large-leaved tea plant of which Linné absolutely 

 ignored the existence, I cannot approve of the idea that Henry's wild 

 tea plants from Yun-nan, as well as the I-bang tea plant and Fortune's 

 Thea viridis, jointly with ail the material accumulated under that name by 

 earlier authors, — that ail those forms should stand in the closest possible 

 relatioiiship to the remarkable large-sized tea plant of Manipur and Lushai. 

 It cannot be denied that China plants may hâve characters in common with 

 Indian and Burmese races (namely, bullate surface and prolonged apex, as in 

 Lettsom's picture 1772), and I am prepared even to admit that such common 

 features may be constituted in identical or similar gènes, — but //, as /i/o, we 

 see tlïe problem of tea varieties essentially as a problem of classification, 

 we will lay more stress on very conspicuous characters, and we shall 

 agrée in the opinion that inevitably we must lose ail review over the 

 varions forms, and ail connection between them will be hopelessly confused, 

 when we classify ail the small-leaved samples quoted above in one map 

 with the Indian types. In a number of instances (not only with référence 

 to var. viridis) I was utterly unable to make out upon what characters were 

 founded Sir Watt's déterminations of herbarium samples. It is tolerably 

 easy to label a given spécimen as „Camellia theifera var. assamica", or 

 „var. sinensis", and (provided the growing place is known!) even „var. 

 burmensis"; but I think it is hardly possible to pick out var. viridis 'Watt 

 from the numerous small-leaved forms with any certainty. 



Ad Ib.— It is a very remarkable fact that var. /as/oca/yx — or to use the 

 dénomination that is entitled to priority, var. pubescens Pierre — has been 



