8AXIFRAQACEM. 407 



cupulate receptacle, the glanduliferous scales alternating with the 

 stamens, the stigmatic lobes superposed to the inultiovulate pla- 

 centas, and the organization of the seeds. 



V. Francoe/e. — A series formed of two genera, one with regular 

 (Francoa), the other with irregular flowers (Tetilla), both comprising 

 scapigerous herbs, with iso- or diplostemonous tetramerous flowers. 

 Carpels as many as petals, cohering into an almost completely 

 superior ovary, with four complete or incomplete cells. Made by 

 several authors 1 a distinct order, linking Crassidacece, Saxifragacece, 

 and Cephalotece ; according to others 2 allied to Ericacece and Pirolece. 



VI. Hydrangea. — Frutescent and arborescent plants. Leaves 

 generally opposite, 3 simple, exstipulate. Petals often valvate. Stamens 

 usually epigynous, diplostemonous or indefinite. Eeceptacle always 

 concave. Ovary completely or partially inferior, with 3-5 cells, com- 

 plete or more frequently incomplete. Hydrangea, alone known to 

 A. L. de Jussieu, was by him placed in the Genera Saxifragaceis affinia. 

 De Candolle 4 made Hydrangea a tribe of Saxifragacece. Lindley 5 

 raised this to a distinct order, adding Bauera. The other genera, all 

 closely analogous to Hydrangea, are of comparatively recent creation, 

 viz., Broussaisia, 6 Cardiandra and Platy crater," and Pileoslegia. 8 

 Dichroa of Loureiro 9 (synonymous with Adamia) dates from the end 

 of last century. 



VII. Philadelphe/e. — Philadeljohus, placed by A. L. de Jussieu 

 in Myrtacece, 10 while Beidzia was relegated to the Genera incertce 

 sedis, 11 was made by Don 12 in 1826 the type of a distinct order, 



doubtfully of an alliance with Baniincirfacece. 2 Don, in Fdinb. N. Phil. Jotirn., Oct., 1828 



AdanSON classed it with other most heterogeneous (Galacineas).—\}C, Prodr., vii. 777. — Lindl., 



o-enera referred to the family Cisti. Jussietj, Veg. Kingd., 451. — " Francoacece are herbaceous 



with many more recent authors (De Candolle, Stachyurece. Stachyurus quite recals Ribes in 



Bbogniart, A. Gray, &c), place it in Drose- habit, Francoa in flower." (Ao., Theor. Syst., 



racece. Endlicheb created an order for it, 152, t. 12, fig. 16.) 



close by Droseracece, a view which Payee held 3 Alternate in Cardiandra, sometimes sub- 

 confirmed by observation on the evolution of the verticillate in Broussaisia. 

 flower. Batsch, Don, Rosper, & Lindley put 4 Prodr., iv. (1830), 13, trib. 4. 

 it in Hyperuinece, or at least close by. Bakt- 5 Veg. Kingd. (1S46), 567, ord. 215. 

 UNO holds it near to Tamariscinece, and 6 Gatjdich., Toy. Freycin., Bot. (1826). 

 Reichenbach to GentianecB. J. E. Smith, R. 7 g IEB . & Zucc, Fl. Jap. (1835). 

 Beo\vn, Lindley (at first), Royle & Planchon 8 In Journ. Linn. Soc, ii. (1857). 

 have thought Parnassia allied to Saxifragacece." ° Fl. Cochinck. (1790). 

 (J. G. Agaedh. Theor. Syst. PL, 80). This 10 Gen. (1789), 325. 

 author himself considers Parnassiece as perfected " Gen., 431. 



Podostemece, leading up to the symmetrical 12 In Fdinb. New Philos. Journ., i. (1826;, 



Lentibulariece. 133. 

 1 Endl., Gen., 812. 



