as distinguished from the Symplocaccfc. 141 



Halesia, the fruit becomes at length wholly infeviov, and crowned 

 by the unchanged toothed margin of the previously inferior, now 

 become superioi*, and adnate calyx*. In other cases in that 

 family (in Olax for example), the calyx enlarges into a tubular 

 form quite free from the fruit which it encloses, as in Stngilia. 

 There are many other points of structure which show the evident 

 relationship of the Styraceoi to the Olacacece. 



Dr. Asa Gray, in his ' Notes on Vavcea/ very justly lays great 

 stress upon the affinities of the Sfi/rncece and MeUacece, and ex- 

 presses his surprise that this fact had not occurred to me. I had, 

 however, clearly implied it by proposing to place the former 

 order close to the Olacaceoi and Humiriacece, among the Ciono- 

 spermece, in which class, if adopted, the MeUacece would occupy a 

 prominent position ; but, notwithstanding all that has been urged 

 by the learned American Professor on this point, to which I readily 

 accede, I still remain of opinion that the proximate alliance of 

 the Styracea is with the Olacacete and Humiriacece rather than 

 with the MeUacece, In the latter family, the leaves are mostly 

 of considerable length and pinnate ; the filaments of the stamens 

 are united into a compact tube, in the mouth of which the 

 anthers are usually quite free and sessile, and where the staminal 

 tube is toothed or lobcd, the anthers commonly alternate with 

 the lobes ; the seeds, too, are generally arillate, and frequently 

 without albumen, with laj-ge fleshy cotyledons, within which the 

 small radicle is retracted : the plants of this family are bitter 

 and astringent, — characters all much at variance with the ^ty~ 

 racece. On the other hand, in Olacacece and Humiriacece, the 

 leaves are simple, the stamens quite free, or only laxly aggluti- 

 nated at base : in the latter family, the filaments are broader and 

 longer than the anther-cells, which are distinct and separated 

 by an interval, and dorsally attached by their whole length to 

 the filament ; the seeds arc albuminous, with a terete radicle, — 

 characters more in conformity with Styracece. The juice of 

 Humirium halsandferum has the same smell and balsamic pro- 

 perties as that of Styrax, from which it is scarcely distinguish- 

 able. The affinity of the Styracece to the Humiriacece was di- 

 stinctly recognized by Prof. Von Martins when he established 

 the latter family f, and was afterwards confirmed by the opinion 

 of Mr. Bcntham|. The character which serves more than any 

 other to establish the relationship of the Styracece with the three 

 families just mentioned, is the imperfect union of the carpels, 

 owing to which the ovary has its dissepiments always more or 

 less incomplete, and therefore it is unilocular in the summit, so 



* Blume, Mus. Bot. Lugd. Bat. i. 251. pi. AG. 

 t Nov. Gen. et Sp. Bras. ii. 142. 

 X Kew Journ. Bot. v. 97. 



