402 Mr. J. Miei's on the Styracefe, 



so that this character aloue ceases to form a distinction between 

 Halesia and Pterostyrax *. 



3. Halesia parviflora, Mich. Fl. Bor. Anier. ii. 40 ; Lindl. Bot. 

 Reg. tab. 952 ; A. DC. Prodr. viii. 270.— In Florida et 

 Georgia. 



7. Pterostyrax. 



We have no knowledge of this genus beyond the description 

 of Zuccarini in Sieber's ' Flora Japonica/ where the typical 

 species is figured. Having seen a specimen in flower in Sir 

 William Hooker's herbarium, I am enabled to s])eak of it with 

 more confidence. The character, as given by Zuccarini, agrees 

 well with the structure of the flower in Halesia, differing in no 

 respect except in the pentamerous disposition of its parts ; but 

 even that distinction T have shown to be of no value, because 

 we sometimes meet with five petals in Halesia, while in Ptero- 

 styrax corymhosa I have found some of the flowers with only four 

 teeth in the calyx, with four petals and eight stamens, all re- 

 sembling in size, shape, and structure those of Halesia, the ovary 

 being at the same time of similar form, and in like manner 4- 

 locellate at its base ; indeed I can perceive no difference what- 

 ever in the structure of the tetramerous flowers of Pterostyrax 

 and those of Halesia. There is, however, a considerable disparity 

 in their respective habits ; and this appears to constitute the 

 principal generic distinction. Zuccarini describes the calyx as 

 having five nervures between the lobes ; but he omits all mention 

 of the carinate nervures, which, proceeding from the points of the 

 teeth, run down decurrently along the pedicel, exactly as in 

 Halesia ; in the latter genus we find the same intermediate 

 nervures as are ascribed to Pterostyrax. The petals are spathu- 

 lately oblong, membranaceous, and free to the base, as in Halesia 

 diptera ; and I find their estivation to be quincuncially imbricate, 

 not sinistrorsely convolute, as stated. The stamens are quite 

 similar in form to those of the species last mentioned, the fila- 

 ments being broad, flat, membranaceous, and pubescent, slightly 

 cohering together at the base, and even there being separable by 

 the slightest force : the anthers are a very little broader than the 

 filaments, the parallel cells separated by an interval, and dorsally 

 adnate to the filaments, not affixed to them by their base, as they 

 are desci'ibed to be. As in Halesia, the ovary is half- superior, 

 and in like manner has a central placenta rising to the middle, 

 leaving the upper moiety 1-locular, its lower portion being 

 divided into four or five divisions, which, branching from the 



* Analytical figures, showing the floral and car])ological structure in 

 these two species, will be given in the ' Contributions,' plate 31. 



