338 DR. SEEMANN'S synopsis OF THE 



generic character of Thea : for instance, the words " Perianthium 6-partitum, 



planum. . . ." are altered into "Per. 5-partitum;" the number "6" assigned to the 

 petals is struck out ; to the description of the stamens is added " filamenta corollaj 

 basi inserta," &c. ; so that, if he had brought out a seventh edition, these corrections 

 would have been introduced, and they would have tended to improve the limits of the two 

 genera. But even giving Linnaeus credit for all these intended improvements, his generic 

 characters of the two contain little that either Camellia and Thea do not share with other 

 Ternstroemiacece, or that more ample materials have shown to be correct. There are, how- 

 ever, two characters upon which he relies to distinguish these genera which have never 

 been shaken — the calyx of Camellia is termed polyphyllous and deciduous, whilst that of 

 Thea is described with a definite number of persistent sepals. Morphologists may possibly 

 object to attach much importance to them, as the calyx of Thea is clad with deciduous 

 bracts, and what is termed a polyphyllous calyx in Camellia may also be explained as a 

 simple calyx surrounded by bracts. This view of the subject would reduce the whole 

 difference to one of time, — the sepals remaining longer attached to the plant in Thea than 

 in Camellia. Systematic botanists, however, are compelled to take a practical as well as 

 a theoretical view of such questions, and that is the course here pursued. 



DeCandoIle, who wrote in 1824, and who looked upon Camellia and Thea as forming a 

 distinct natural order, chiefly differing from Ternstroemiacece by the absence of albumen, 

 thus defines (Prodr. i. pp. 529, 530) the genera in question : — 



Camellia, Linn. 

 Calyx imbricatus, nempe bracteis sepalisve non- 

 nullis accessoriis cinctus. Stamina basi poly- 

 adelpha aut monadelpha. Antherse ellipsoideEe. 

 Capsula valvis medio septiferis, axim triquetrum 

 liberum post dehiscentiam relinquentibus. 



Thea, Kaempf. 

 Calyx 5-6-sepalus. Petala 6-9 ima basi subco- 

 hajrentia 2-3-serialia. Stamina basi sublibera. 

 Antherse subrotundfe. Capsula .3-coccn, septis 

 valvaribus, nempe a valvularum marginibus in- 

 troflexis formatis. 



DeCandoIle places under Camellia one species which has no right to be there, viz. Camel- 

 lia axillaris, Roxb. { = Folyspora axillaris, Sweet), while under Thea he ranges Loureiro's 

 Tliea oleosa, which I hold to be the old oil-seeded Camellia Sasanqua of Thunberg ; and it 

 is possible that in di'awing up his generic characters the presence of those plants may have 

 influenced him in some degree, compelling us to make some allowance in criticising them. 

 He terms the stamens of Camellia polyadelphous : that term must, I think, be suppressed, 

 as it cannot be applied to those of any true species of that genus, unless we are prepared 

 to apply it to the result of the mechanical rvipture of the tube of the monadelphous stamens, 

 Avitnessed in Camellia Sasanqua when the petals begin to fade, bend back, and become 

 detached from the stamens. In Thea he describes the stamens as nearly free, but the 

 outer series of them is always monadelphous, as in Camellia ; and, following Gsertner, he 

 assumes the cells of the capsule, imlike those of Camellia, to be formed by the edges of the 

 valves being bent inwards. The latter are, however, exactly formed as those of Camellia, 

 the septa being placed on the middle of the valve. DeCandoIle has therefore not discovered 

 any additional character by which Camellia and Thea could be distinguished from each 

 other. 



We now come to W. B. Booth, who in 1830 published an able history of the genera 



f 



