BOTANICAL DIFFERENCE. 323 



as the Bohea (Linn.). Again, De Candolle, though 

 he regards the Cantoniensis as the Thea Bohea, yet 

 considers the Stricta of Aiton (which we have here 

 supposed to be identical) to be the Viridis. In these 

 respects De Candolle adopted the views of Sims. 



Moreover, botanists now agree that no specific 

 difference can be established on the number of 

 petals of polyandrous plants, in consequence of the 

 disposition of the stamens to transform into petals. 

 Then, if the corolla be rejected as a sign of species 

 when its character is not sufficiently permanent, 

 can it be said that the other characters of Loureiro 

 prove any thing more than that the Cochin Chinese 

 plant is one of luxuriant growth, and the Canton 

 a stunted distorted shrub ? " Arbor octo pedes 

 alta ; ramis diffusis," the one ; and " Arbuscula 

 quatuor pedalis, ramosissima, tortuosa," the other. 



Again, if we examine the shape, size, and form 

 of the serratures and colour of the leaf, we shall 

 find them, as it appears to me, from their incon- 

 stancy, equally unsatisfactory signs of species with 

 the corolla. Lettsom describes the leaf of the 

 Yiridis as " latitudine triplo longioribus," and Sims 

 the Bohea, " latitudine duplo longioribus;" which 

 descriptions agree sufficiently well with that of 

 Linnaeus, " Bohea ovalia, Viridis autem oblongiora 

 habet folia." But I have impressions of some 

 infused leaves of a fine Paochong tea in daily use 

 in the year 1823, taken accidentally and without 

 any definite object, which correspond not with the 



Y 2 



