340 DR. SEEMANN'S synopsis OF THE 



discovered in Fyrenarin. On the otlier liand, the structure now discovered in Camellia 

 (a definite number of free stamens surrounded by a series of monadelphous ones) was 

 known to exist in Calpandria of Blume, placed by Endlicher (Genera Plantarum, n. 5546) 

 among tlie " genera dubia " of Ileliacece, and removed to Ternstroemiacece by Choisy, who, 

 however, did not suspect that it was so intimately related to Camellia as to render a 

 generic separation impossible. True, in Camellia the filaments of the monadelphous 

 series of stamens are more free towards their upper extremity than those of Calpandria ; 

 but it must be admitted this is a mere matter of degree ; and no one would like to keep up 

 Calpandria as a sejoarate genus on that account only, especially as it agrees in habit and 

 all other respects with CavielUa. 



The number of styles in Thea I have always found to be three, whilst in Camellia there 

 are normally five, or by abortion four or three. I here use the term ' styles ' for what 

 some authors have called stigmas ; and I justify this application by pointing to Camellia 

 Songkongensis, where the styles are quite free to their very base. There is, besides, 

 another feature, which, though it cannot be admitted into the technical characters of the 

 two genera, serves to distinguish them by their habit ; I mean that in Camellia the 

 flowers are always erect, whilst in Thea they are nodding. Thus, if we sum up the 

 distinguishing peculiarities of the two, we have the following : — 



Camellia, Linn. I Thea, Kaempf. 



Calyx polyphyllus, sepalis deciduis. Stamina in- i Calyx bracteatus, 5-sepalus, sepalis persistentibus. 



teriora duplo petalorum numero. Styli 5 (ab- Stamina interiora petalorum numero Eequalia. 



ortu 4 V. 3).— Flores sessiles, erecti. I Styli 3. — Flores pedunculati, declinati. 



When we apply these characters to the aggregate body of Camellias and TJieas as it now 

 stands, we are enabled to retain all the older typical species under their respective genera. 

 Thus Camellia Japonica of Linnaeus, together wdth C. reticulata, Lindl., C. Sasanqiia, 

 Thunb., and drupifera, Lour., are kept under Camellia, whilst Thea Chinensis, Linn., 

 remains undisturbed under Thea. Only four modern species of Camellia are placed with 

 Thea ; and that is all the change required. A few other species, which have nothing to do 

 either with Camellia or Thea, are refei'red to their proper systematic position : Camellia 

 axillaris, Roxb., is admitted to be the same as J^olyspora axillaris. Sweet ; C. integrifolia, 

 Chois., proves itself to be a Laurinea {Aclinodaphne Chinensis, Nees), according to speci- 

 mens kindly transmitted by Prof. Choisy himself; C. ? Scottlana, Wall., is held to heAndi- 

 nandra dumosa. Jack ; whilst a plant which had been considered to be allied to the Wild 

 Tea of Assam, and had been distributed by Wallich under the name of Freziera ? atte- 

 nuata, I found to be a species of Fyrenaria*. 



I now proceed to give the generic characters of Camellia and Thea, and a synopsis of 



* Pyrenaria attenuata. Seem, in ' Bonplandia,' vii. p. 49 [1859] (name only) ; foliis obovatis acuminatis serratis 

 basi atteuuatis glabris, floribus solitariis pedunculatis, pedunculis medio bracteis lunulatis obtusis alternis instructis, 

 sepalis (5) suboibiculatis obtusis bracteisque ciliolatis, petalis (5) cohEerentibus obovatis v. oblongis obtusis, 3 intC' 

 rioribus paulo majoribus, omnibus dorse sericeis, staniinibus interioribus liberis 5, exterioribus monadelphis nuraerosis, 

 omnibus glabris, sty lis 5 liberis glabris, stigmatibus capitellatis, ovario ovato acuminato 10-sulcato hirsuto, fructu . . 

 K^ms P. masocarpce, Korth. — i^;-eji'era ? attenuata, AVall., an Camellia 1, Wall. Cat. n. 1451! — TAea Assamica 

 affinis sp., Chois. in Mem. Soc. Gen. xiv. i. p. I.tG (1855). — Thea viridis, var. Asxamica ?, Planch, in Herb. Hook. 

 Chois. I. c. p. I7i).—Hai. in Tavoy (Gomez ! in Wall. Cat. n. 1451). 



