ILLUSTRATIONS OF INDIAN BOTANY, 



111 



Remarks on Genera and Species. The old genus Lobelia 1ms, Ly recent wrller«^ been 



Lobelia 



synonyms. To what extent all these genera are well founded I am quite unable to say, beinff 

 acquainted with only two or three of them. All the Indian species, except two, are gtill placed 

 m the original genus, and I have not specimens of either of the off-bhoots for examination. The 

 most remarkable among the Peninsular species is undoubtedly L. cxcdsa, so abundant on 



Ne 



th« 



feet, the lower half quite naked, the middle thickly covered with leaves, and ending in a long 



denge, tapering spike of pale yellowish, or slightly variegated flowers, but so intermixed with 



leafy bracts, as to be almost concealed by them. In Ceylon there is a nearly allied species 



Moon (Wight's Ic. No 



ifolia 



L. trichandra (R. W. Icon. No. 117J) 



the Neilgherries, below Sispara, flowering in February and March. 



EXPLANATION OF PLATE 135. 



Lobelia ^Nicoiianifolia 

 1. Detached flower. 



Flowering branch. 



2. The same, the corolla removed, showing the epig- 

 ynous insertion of the stamens, the anthers forcibly 

 separated and each bearded at the apex. 



3. Ovary cut vertically. 



4. cut transversely. 



5. Mature capsule after dehiscence. 



6. The same cut vertically, showing the position and 

 attachment of the placentae. 



7. The same cut transversely, showing the inflexed 

 placentiferous, carpellary margins. 



8. Half of the capsule divided loculindally as in 

 natural dehiscence, 



9. 10. Upper and under surfaces of the leaves. 



11. Ovary of Lobelia excdsa^ showing that it is nearly 

 free, not adherent to tube of the calyx as in the other. 

 All more or less magnified. 



LXXXIX 



.E 



innocent 



blueing about 25 genera, and probably not fewer than from 450 to 500 species. These are 

 widely distributed over extra-tropical regions, but witliln the tropics are very rare, except in 

 alpine situations. The plains of India only furnish three or four species, and not a single true 



Campanula^ those described under that name being all referable to Wahlenhergia. On the 



^ . /. ^ , , J thmk two Wahlenbergias, hence, as 



Neilgh 



an Indian family, it is one of very secondary importance. Alph. De Candolle, who has published 

 an admirable monograph of the order, thus characterizes it : 



F 



Character of the Order. Calyx usually 5-lobed, occasionally 3-6-8- or 10-lobed, ad- 

 nate to the ovary, the lobes equal Corolla, petals united (gamopetalous), regular, or rarely 

 somewhat irregular, divisions alternate with the lobes of the calyx, valvate in aestivation. Sta- 

 mens 3-5-6-8 or 10, usually equalling, never exceeding the lobes of the corolla, alternate with 

 them, and not adhering to the tube: filaments usually dilated, membranaceous at the base: 

 anthers for the most part free, the cells bursting longitudinally before dehiscence. Pollen grains 



(when dry) 



Ovary inferior, 2-10-celled, in Alerciera 1- 



celled from the incomplete partitions. Ovules numerous, in Merciera 4. Styles more or less 



Stigma naked, sometimes bound with an indusium, 



caducous 



usually branched, the branches equalling the cells of the ovary, erect in the flower-bud, hairy on 

 the back, papillose within, diverging or recurved in the flower. Cells of the ovary, when equal 

 to the lobes of the calyx, either alternate or opposite to them. Capsule dehiscing at the apex or 

 sides, the valves for the most part bearing the partitions more rarely without valves, opening by 

 pores or fissures. 



Herbaceo 



suiFruticose, milky plants, most frequent in teniperate regions. Leaves exstipulate, alternate, or 

 rarely opposite, often dentate. Inflorescence either definate, centrifugal or obscurely indefinate. 



case 



Flowers solitary or 



glomerate, generally pedicelled; seldom involucrate. Corolla usually blue, sometimes yellow 

 or purple. 



M 



