CAMPANULACEOUS AND OLEACEOUS ORDERS. 15 



petals ; but this isomery is only a further continuation of the bi- 

 nary arrangement from the pistil outwards. 



I know of no order with which the Oleaceous group is imme- 

 diately connected beyond the gereral affinity to bicarpellary 

 Gamopetalae, nor am I aware of any ambiguous or connecting 



genus. 



Of the five subordinate groups, the Salvadoraceae are certainly 

 the most distinct, although they have no one constant character 

 which is not to be found in some genus of the other groups, ex- 

 cept perhaps the rudimentary stipules, of which there is, I believe, 

 no trace in any Oleaceae. The glandular scales behind the sta- 

 mens ofDobera, and sometimes of Salvadora, have no existence in 

 Oleaceae, but are not constant in Salvadoraceae. The stamens are 

 always four, as in one or two species of each of three genera of 

 Oleaceae ; the radicle is inferior, as in Jasmineae ; the general re- 

 semblance more with Oleineae. 



Jasmines, as well as Salvadoraceae, are frequently treated as a 

 separate order, the distinctive characters usually given being an 

 increased number in the parts of the corolla, a didymous fruit, 

 exalbuminous seeds, and an inferior radicle. But the increase in 

 the number of corolla-parts occurs also in Schrebera, with the 

 loculicidal capsule and inferior radicle of Syringeae ; the fruit is 

 not didymous in Nyctanthes, which in all other respects is a true 

 Jasminea; there is no albumen in Schrebera among Syringeae, 

 nor in Noronhea among Oleineae ; and in Linociera the albumi- 

 nous and exalbuminous species are sometimes most closely allied 

 in all other respects. There remains the inferior radicle ; but 

 even this character may not be here considered in the same light 

 as in the case of Boragineae and Verbenaceae. The ovules are in 

 all Oleaceae, as far as I have observed, anatropous and laterally 

 attached ; and in the seed the radicle, always next the hilura, be- 

 comes inferior when the seed is erect, superior when it is pendu- 

 lous. On the other hand, the radicle is always superior in Bora- 

 ginese, and inferior in Verbenaceae, whatever be the position of the 

 hilums. The climbing habit of many Jasmina does not extend to 

 Menodora or Nyctanihes, and occurs in Myxopyrum among Oleineae. 

 The dissected leaves are also found in Fraxinus, and sometimes 

 in Schrebera and Syringa. 



The three remaining subordinate groups or tribes are distin- 

 guished chiefly by their fruit — capsular and loculicidally dehis- 

 cent in Syringeae, samaroid, winged and indehiscent in Fraxineae, 



