774 



APIACEiE. 



[Epigynous Exogens. 



experienced eye, so much are they at variance with the usual structure of Umbellifers 



In all these cases, however, the very pecuhar 

 condition of the flower and fruit is abundantly 

 sufficient to mark the Order. Indeed we have 

 no knowledge of any one group so entirely free 

 from deviations from the typical structure, ex- 



Fig. DXII. 



Fig. DXIII. 



cept m accidental monsters. Of these I once found an instance, at Bm-nham Priory, 

 near Maidenhead, m which the calyx was detached from the ovary, which had become 

 superior, the calyx surromiding it 'loosely like a 5-toothed ribbed cup. This is quite 

 mconsistent with the theory of Schykoffsky, who assumes that m Umbellifers the calyx 

 proceeds really from the same point as the styles.— Bot. Reg. 1841. Misc. 35. 



_ It is also stated that in some accidental cases 3 carpels have been found. In Eryn- 

 gium and some Bupleurums the leaves consist of nothuag but petiole, and thus present 

 the appearance of Endogens. Among the more remarkable facts connected with the 

 structm^e of the fruit are, 1 , the separation of it when ripe into 2 carpels or mericarps, 

 adhering to a stylopod or forked placenta, eventually exterior to the carpels themselves, 

 although in the begimiing it must have been included between their confluent margins, 

 between which it rose till near the summit of the ca\dty, when it turned inwards to 

 bear the sohtary ov-ules ; and 2, the presence in the pericarp of fistular passages filled 

 with oil ; the latter are no doubt analogous to the cysts of Orange and other leaves, and 

 to the glands of Labiates and some Composites, but they are remarkable for a unifomiity 

 in position and number, which, although not absolute, is nevertheless very diff'erent 

 from the indefinite nature of common cysts. 



UmbelUfei's differ from Ivyworts in then* seed adhering to the pericarp, in their 

 imbricated corolla, and in the two divisions into which their di'y fruit always resolves 

 itself eventually. Ivyworts, on the contrary, have a loose seed, a valvate corolla, and 

 more divisions of their succulent fruit than two. The genus Horsfieldia, however, forms 

 a complete transition, having the valvate corolla of Ivyworts and their peculiar habit, 

 with the dimerous dry fruit and adherent seed of Umbellifers. As to then- other 

 affinities it may be remarked, that they completely represent in the epigynous sub-class 

 the Crowfoots among hypogynous Exogens ; some Thalictrums mdeed would make 

 pretty good UmbelUfers, if their calyx adhered to the side of the ovary. They ap- 

 proach Stellates in their didymous inferior fruit and copious albumen, but they are 

 universally polypetalous. With Saxifrages Umbellifers agree in habit, if Hydi'ocotyle 

 is compared with Chrysosplenium, and if the sheathing and divided leaves of the two 

 Orders are considered. To Cranesbills De CandoUe remarks that UmbelUfers are 

 alhed, m consequence of the cohesion of the carpels around a woody axis, and of tlie 

 umbellate flowers which grow opposite the leaves, and also because the affinity of 

 C ranesbills t o Vines, and of the latter to Ivyworts, is not to be doubted. The resem- 



Fig. DXII.— Leucolaena rotundifolia.— Hooftcr. 

 Fig DXIII.— Bolax glebaria.— /JooSer. 



