ROSACEA. 431 



nearly always solitary. Style <^ynobasic. Ovules geminate collateral, 

 ascending, micropyle inferior, looking towards the insertion of style. 

 Stem woody. Leaves simple.' 



Affinities. — Many authors have pointed out the relations of the 

 Bosacece to those Poly earpiece near which we now place them. The 

 CalycanthefE have even been placed in the order Bosacece, and we know 

 how close are their affinities with Magnoliaeecs, especially with Illiciece, 

 from which they only differ in the form of the receptacle. Now as 

 the receptacle is altogether that of the Roses, the only character 

 separating Calycaniltece from Rosaeea, is the arrangement of the pieces 

 of the androceum, spiral in the former, verticillate in the latter. 

 But this difficulty is really of the less fundamental importance, as in 

 Bammeulaee^ we have seen genera with curviseriate, and others with 

 verticillate stamens, united in one and the same order. This last 

 order has been violently removed from Rosacea by an over-strict 

 application of A. L. de Jiissieu's principles as regards the value of 

 the insertion, the direct consequence of the configuration of the 

 receptacle. The Rosacea once relegated in Periyynce, far from the 

 hypogynous orders of which Ranuneidacea are the commonest type, 

 people naturally overlooked the striking identity of all the other 

 parts which leads the vulgar instinctively to put white or yellow 

 flowered Crowfoots [Bo idons-d' argent or d'or) in the same category with 

 Potentils, whose flowers are of the same colour. No doubt hypogyny 

 gives a distinct general character to Rnnuneidacea, just as perigyny 

 does to Rosacea ; but we must bear in mind that the perigyny is 

 nearly lost in Stylobasium, as in many of the Frayariea, while on 

 the other side there are Ranunculacea with a slightly perigynous 



* It is impossible to make any general study (1841), Suppl., ii. t. 25), 'Rosa (Met^x, An. 



of the anatomical structure of the Rosacea. Unci. Fhys. d. Gew. (1836), t, iii. 11). — See also 



Some of the trees of this order are of the num- Oliver, The Struct, of the Stem in Dicot., 12. 



ber of those which have served to establish the The Chrt/sohalaneri' have been little studied 



generally admitted type of the ordinary struc- from a histological point of view (see II. MoiiL., 



ture of the stem in Dicotyledons. Such are Rot. Zeit. (1861), 211, and Wicke, 97). It is 



especially the genera Primus and Pyrus (MiiiB., uncertain whether the singular plant called 



Mem. Mus., xvi. (1828), 29, 30 ; — Link., Icon. Canto or Canta, from the Antilles, studied by 



Sel, fasc, i. vi. 1-3; viii. 3-5;— H. Mohl., Ckueoek (i?o^. Ze//. (1857), 281, 29S), reuiark- 



Bot. Zeit. (1855), 879; — Schacht. Der Baum, able for its parenchyma intersected by bands of 



195 ; — WiGAND, Uber die Organis. d. PJlanz., a peculiar cellular tissue, and the deposits of 



Pringhs. Jahrl)., iii. 115), Ruhus (KiES., cellular tissue in its stem, it is uncertain, we say, 



Mem. sur I'Org. des PI. (1814), t. 16; — whether this is really a Chrysobalanad. 

 ScHULTZ (C. H.), Die Cycl., Nov. Act., xviii. 



