as distinguished from the Symplocacese. 897 



ferrugineo-tomentosis, pilis fasciculatis ; folia alterna, ohlongo^ 

 elliptica, Integra, utrinque ohtusa, coiiacea, superne glabra; 

 racemi axillares, simplices, foliis duplo longiores ; pedicelli «/- 

 ternij calyce breviores ; iloves parvi, 



1. Foveolaria ferruginea, E.&P. Syst. 100; DC.Prodr. viii. 272; 

 Deless. Icon. Sel. v. 19. tab. 43. Tremanthus fcrrugineus, 

 Pers. Erich, i. 467. Strigilia raceinosa, DC, Prodr, i. 621 

 (non Cav.). — Peruvia. 



6. Halesia. 



I have already given full details of the structure of the ovary 

 in this genus, of the curious mode of development of its ovules, 

 and the growth of its fruit and seed, all of which show clearly, 

 notwithstanding these anomalous appearances, that it belongs 

 truly to Stijrace(2. In habit, the species bear great external re- 

 semblance to those of Styrax ; the petals are of the same form, 

 colour, and size, the aestivation of the corolla is similar, and in 

 their stamens there is a remarkable resemblance : but the flowers 

 are not racemose, as in that genus ; on the contrary, they grow 

 upon slender, drooping, solitary peduncles, which are few, and 

 fasciculated at the ends of the terminal branchlets, appearing in 

 the axils of the fallen leaves of the previous year's growth, before 

 the new leaves sprout, so that the branchlets then bear the sem- 

 blance of racemes ; but before the fall of the corolla, other young 

 branchlets grow out of the axils with great rapidity, to a con- 

 siderable length, producing fresh leaves in abundance, which 

 completely destroy the racemose appearance of the branchlets. 

 The peduncles are longer than the flowers, which resemble those 

 of Styrax officinalis. In their structure the anthers quite con- 

 form to those of Styrax and Strigilia, the two parallel linear 

 lobes being separated by a considerable interval, quite adnate 

 for their entire length upon a more or less broad ligular fila- 

 ment, very thin and membranous in texture, and bursting in- 

 wardly by a longitudinal fissure : the filaments continue broad 

 and compressed to their base, where they are laxly connate for 

 a short distance, and at the same time they slightly agglutinate 

 together the bottom of the claws of the petals — all being easily 

 separated without any laceration of the parts ; indeed before the 

 flower fades they become detached from one another of their 

 own accord ; it is therefore incorrect to describe the corolla as 

 monopetalous, and the stamens as monadelphous. I have ob- 

 served that in H. tetraptera the style is simple throughout its 

 entire length; but in H. diptera it most frequently (but not 

 always) divides into three distinct thread-like portions for a 

 considerable distance from the summit. There is generally 



