690 CAMPANULACE^. [Epigynous Exogens. 



nulas the cells of the ovary are equal m number to the stamens and the divisions of the 

 corolla and calyx, which points out the natural symmetry of the flower. In Lobehads 

 abortion is more frequent. In both groups the innermost organs are abortive more 

 frequently than the outennost. Thus, the number of cells is often smaller (never 

 greater) than that of the stamens ; the number of stamens is sometimes smaller (but 

 never larger) than that of the lobes of the coroUa ; and the same is time of the lobes 

 of the corolla with respect to the calyx. Finally, Lobehads have sometimes a corolla 

 of a fine bright-red, a colom' unknown among Bellworts ; nine-tenths of the species 

 of the latter have blue flowers ; and those in which the colom' varies, and into which a 

 little red enters (as Canarina), are far from haAing the brilliancy of Lobelia cardinalis 

 for instance. After Lobehads, the Natural Orders with which Bellworts have 

 the most relation are, no doubt, Goodeniads and Styleworts, which fonned part of the 

 Campanulee of Jussieu. The regular corolla of Bellworts distmguishes them, at first 

 sight, from both those Orders, as well as from Lobehads. Besides, Campanulas have 

 not the fringed indusium which terminates the style of Goodeniads and smTOunds their 

 stigma. Although this organisation approaches that of Lobehads, and so of Bellworts, 

 it is not less true that it affbi'ds an important mark of distinction, and that it is con- 

 nected with essential differences in the mode of fecundation. Bi'own has also remarked, 

 that the corolla of Goodeniads is sometimes polypetalous, which it never is in Bellworts 

 or Lobehads ; that the aestivation of the corolla is induphcate, not valvate ; that its 

 principal vems are latei'al, or alternate with the lobes, as in Composites ; that in the 

 species of Goodeniads Av-ith dehiscent fruit, the dehiscence is usually septicidal, while in 

 the two other groups it is always loculicidal ; finally, that Goodeniads have not the 

 milky juice that characterises Bellworts and Lobehads." Notwithstanduig the poly- 

 spermous fruit and different inflorescence, this Order approaches very closely to Com- 

 posites ; the milky juice is the same as that of the tribe caUed Cichoracese ; the species 

 have, in many cases, the flowers crowded m heads ; the stigma is similar to that of 

 many Composites ; there are the same collecting hairs on the style, in both cases 

 intended to clear out the pollen from the cells of the anthers ; and, finally, the habit is 

 very hke. These collecting hau'S, wliich clothe the style of Bellworts in a most remark- 

 able manner, arranging themselves in lines haAdug a du'ect relation to the number and 

 position of the anthers, have been the subject of special examination 

 by several observers, especially by Adolphe Brongniart. Tliis Bota- 

 nist ascertained that such hau'S are not, as had been supposed, 

 deciduous, but that they are retractile, hke the hah's of certain 

 annelides or the tentacula of snails. It appears that, at the time of 

 the expansion of the flower, the hau*s, which had previously projected 

 and swept out the pollen from the anthers, are drawn back into 

 certain caAdties lymg at theu" base, the upper half sheathing itself 

 in the lower half as it is by degi'ees \rithdi'awu. M. Brongniart is of 

 opinion that there is no ground for supposing that tliis singular 

 phenomenon is connected \rith the fertilising process. (See Ann. 

 des Sc. Nat. 2 Se7: 12. 244). But Mr. Hassall disputes this statement, 

 which he declares is "wholly opposed to the result of his investiga- 

 tions." — Ann. Nat. Hist. Aiii. 86. 



It has been remarked in the Botanical Register (1842, t. 3.), that 

 the genus Glossocomia brmgs the Orders of Nightshades and Bell- 

 Fig. CCCCLXIV. y^^orts mto close contact. 



With respect to the singular genus Sphenoclea, erected into an Order by Mar- 

 tins, although it cannot be regarded as a genuine species of Bellwort, because 

 of the absence of collecting hau's from its styles, the roimd sub-sessile anthers, 

 the stamens distinctly inserted upon the corolla, and the pecuhar habit of the only 

 knoM-n species, yet it seems to have more affinity to this Order than to any other, and 

 may very well be stationed at the end, as a genus waitmg for the discovery of com- 

 panions which may be better suited to indicate its time station. 



Chiefly natives of the north of Asia, Em'ope, and North America, and scarcely known 

 in the hot regions of the world. In the meadows, fields, and forests of the countries 

 they inhabit, they constitute the most striking ornament. Some cm-ious species are 

 found m the Canaries, St. Helena, and Juan Fernandez. Alphonse De Caudolle 

 remarks, that "it is within 36° and 47° N. lat. that m our hemisphere the greatest 

 number of species is found ; the cham of the Alps, Italy, Greece, Caucasus, the Altai 

 range, are their true country. In whatever direction we leave these hmits, the number 

 of species rapidly decreases. In the southern hemisphere, the Cape of Good Hope 

 (lat. 34° S.) is another centre of habitation, containing not fewer than 63 species. This 



Fig. CCCCLXIV.— Stamen and pistil of Campanula Medium. 



