186 A CIIINKSE FONTANESIA. 



indebted to liim for excellent specimens of the shrub referred to 

 b}^ Mr. Hemsle}'. An examination of this proves unequivocally 

 that it is in all respects a true Fontancsia, of which I subjoin 

 a diagnosis. 



FoNTANESiA cHiNENsis, s]). tiov. — Frutcx V. arbuscula, ramis 

 cortice ciuereo obductis, ramulis erectis glaberrimis, foliis glaber- 

 rimis lanceolatis sensim acuminatis integerrimis x^enniveniis subtus 

 pallidioribus li-2^ poll, longis infra medium 6-9 lin. latis petiolo 

 lineali, floribus in racemos axillares et terminales folio breviores 

 V. fere aaquilongos dispositis subpolygamis ahis scilicet lierma- 

 j^hroditis aliis masculis ovario nunquam maturescenti, pedicellis 

 flore subffiquilongis, calyce brevi inasqualiter 4-lobo, petalis 

 oblongis basi per paria connexis, antheris petala subdimidio 

 excedentibus ipso filamento crassiusculo 5-6-plo longioribus, ovario 

 florum fertilium calyci squilongo orbiculari stylo eo triplo longiore 

 coronato stigmate bifido, loculis semper* uniovulatis, fiorum 

 sterilium cito ad 5-Q» lin. longitudinis excrescente oblongo fere ad 

 medium usque in lobos 2 angustos acutos fisso loculis ssepissime 

 vacuis nunc ovulum cassum gerentibus, samara quadrato-oblonga 

 utrinque retusa 4 lin. longa. 



In colhbus Feng-wang-shan, provincige Kiang-su, florif. m. 

 Maio, frf. mm. Octobri et Novembri, 1877-8, collegit mecum- 

 que communicavit amicissimus F. B. Forbes. (Herb, propr. 

 n. 20725.) 



The hills where this plant grows, I learn from Mr. Forbes, are 

 those nearest to Shanghae, lying about twenty miles to the S.W. 

 of that city. The highest is not more than 250 feet above the sea- 

 level ; and he tells me there is no doubt that they once belonged 

 to the Chusan archipelago, of which another small group is to be 

 found silted up at Clia-pu, on the coast S.E. of Shanghae, where 

 the sea still washes then- base. 



In the herbarium — I have seen neither shrub alive — this species 

 has the most striking resemblance to the Syrian F. philUjrceoides, 

 Labill., the only other one known; so great indeed that, without 

 examination, the two might readily be supposed to be identical. 

 But it differs by its more truly lanceolate rather than oblong or 

 oblong-lanceolate leaves, its more fully developed inflorescence, 

 the rather shorter pedicels, short filaments, uniovulate ovary-cells, 

 longer style, and especially by its subpolygamous flowers, — perfect 

 and imperfect ones intermingled, — with the curious rapidly growing 

 abortive ovary of the sterile foi"m. The occurrence of these sterile 

 flowers is interesting, as showing apparently that the aflinity of the 

 genus is rather with Fraxinus, next which it is placed by Bentham 

 and Hooker in the " Genera," than with the Si/riiKjew, where 

 DeCandolle stationed it. But it must be acknowledged, I think, 

 that the genera in this order cannot be marshalled into very well- 

 marked tril)es, and Boissier (excluding Jusmineie) admits only two, 

 Oleineoi with a fleshy fruit, and Lilacece with a diy one.f 



* Ex iteratis am. Forbesii observationibus. 

 f Fl. Orient, iv., 58. 



