7S 



THE NATURALIST. 



I^OTES ON SOME FORMS OF CRAT^GUK- 



C, P. HOBKIRK. 



( Concluded from page Qb.) 



From tiie want of authentic specimens of tlie various authors, ancE 

 having only their written descriptions to depend upon, I find it almost 

 impossible to settle the synonymy of the forms described, no 

 one of them agreeing altogether with any other. Linneus' oxyacantha 

 clearly includes all the forms, and must either be thrown aside altogether 

 or otherwise affixed to one of the forms and be more clearly defined. 

 Then arises the question which is the true oxyacantJioides of Thuillier ? 

 This I have endeavoured in vain to ascertain. His principal character 

 seems to be that the leaves are glabrous and not serrate — but not one of the 

 specimens I have, thus labelled, have these characters ; most of them have 

 more or less serrate leaves, and are also more or less hairy on both sides ; 

 one form from near Lyon labelled by M. A. Jordan, has the leaves plain^ 

 but hairy on the midrib and principal veins on both surfaces, whilst Mr. 

 Boswell Syme's specimens from Hampstead Heath have finely serrate leaves 

 which are also hairy on both sides. Boreau describes Thuillier's oxyacan- 

 tJioides as having "inciso-dentate lobes, glabrous," (ante p. 12) with which a 

 specimen from Dered-el-Keber near Blidab, Algeria, agrees well, as also in its 

 other characters, and on this account I should feel inclined to adopt his 

 diagnosis as the typical oxyacantJioides and include some of the other forms 

 as varieties. 



Jacquin's description of monogyna is more precise and particular than 

 Thuillier's of oxyacantJioides, and on this account I would adopt his diagnosis, 

 with some slight modifications, as that of the typical monogyna. Two forms 

 which have hitherto been included under this name may safely be separated 

 as distinct species, and some others may be included as varieties. 



After a careful examination and comparison of the forms in my posses- 

 sion, I would then propose the following division into species and varieties of 

 the forms hitherto included under the above two names. 



i. C. OxYACANTHOIDES. Thuil. 



C. spinosa, Miis obovatis, trilobatis, serratis, utrinque glabris, lobis nervisque con- 

 vergentibus, pedunculis glabris, floribus corymbosis 2-3-gynis calycibus laciniisque 

 glabris, his piano -expansus, acutiusculis reflexis, fructibus globosis, nucleis 2-3 includen- 



tibiis, 



