320 



NATURAL niSTOBY OF PLANTS. 



the presence ol' an abortive ovule beside the sterile one ; a cha- 

 racter almost constant, it is true, in Calycaiithus, but not invariably 

 found in llortonia, and never observed in any of the other genera of 

 llortoiiU'd'. 



3. Last of all come those characters which ver}^ frequently vary, 

 their absence and presence being about equally frequent; so that 

 none are sufficient to do more than distinguish different genera, or 

 dirt'erent sections of a single genus. These are — the presence of 

 glandular dots, and the resulting aroma ; the conformation of the 

 hairs covering certain organs, especially the leaves -^ the dehiscence 

 of the anthers in straight or curved lines, or by valves ; their aspect ; 

 tlie presence or absence of glands at the base of the filaments ; the 

 way the floral receptacle opens to free the pollen in the male flowers 

 or the carpels in the fruit, whether by longitudinal clefts, by a trans- 

 verse circular solution of continuity, or by the simple dilatation 

 of its superior orifice ; the consistency of the different parts of 

 the fruit — the indusium and the true pericarps ;" and finally, the 

 absolute direction of the ovules and seeds, whether ascending or 

 descending/ 



From a histological point of view, the vegetative organs in this 

 order are always very uniform/ The stems and branches are cylin- 

 drical or slightly quadrangular. In the aromatic species the bark is 

 always the part that is richest in odoriferous matter, and is very 

 often the only portion that contains any. Usually, as in Peunms, 

 Hortonia, certain species of Mollinedia, and the Atherospermea, the 



' See p. 300, note 2, and p. 322. 



- Following A. L, De Jussieu, Tulasne has 

 based two of his tribes on this character; Athe- 

 rospermem {Ach^niophorecB) with dry fruits, 

 and Monimiacem (DrupacecB) with the pericarps 

 partly fleshy. We luive s-how n {Adansonia, ix. 125) 

 that there are numerous transitions between the 

 drupes and the achenes ; that the CaJycanthece 

 originally possess drupes with thin pericarps, 

 that the fruits of ISijmruna are, so to speak, half- 

 drupes, and that certain Alherospermem have 

 rather caryopsids than achenes. Besides Tulasxe 

 has clearly i)erceived {Mon., 425) that Jlorioiiia 

 affords a transition between the true Monimiacem 

 and the Atheroxpermem in the characters of its 

 fruits. 



3 A. Dk Caxdoi.le {op. rl(., 641) has used 

 this character to distinguish the five tribes he 

 admits in this order. His Tamhottrisseee, Moni- 

 mietr an! Uedticnrypct would only include genera 



with pendulous ovules, while the plants of his 

 AiherospermecB and SiparunecE would have erect 

 ovules. But this last tribe is evidently hetero- 

 geneous, including Siparuna, whose ovule is 

 ascending, and Palmeria, in which it is descend- 

 ing. We have said [Adansonia, ix. 130) that 

 Palmeria is hardly generically distinct from 

 Monimia. Le Maout & Decaisne [Traite Gen. 

 de Hot., 517) have asserted that the absolute 

 direction of the ovule brings about a peculiar in- 

 sertion for the style. " Ovule . . . sometimes 

 pendulous, and then style terminal ; sometimes 

 erect, and then style lateral or basilar." The 

 facts are contrary to this law; out of three 

 genera, with erect ovules, two liave the style not 

 lateral, but thoroughly terminal, viz., Siparuna 

 and Afherosperma (including Laurelia). 



•• TuL., Men., 282, iv.— Oliv., the Slrnct. of 

 tlw Stem in Dicot., 30. 



