ILLUSTRATIONS OF INDIAN BOTANY. 129 



Flowers regular. Stamens equalling the lobes of tl>e corolla. Ovary composed of 

 2 carpels anterior and posterior, 2- or spuriously 4-celled. Cells 2- or 1-seeded. 

 Fruit drupaceous or composed of 4 achaenia. Seed exalbuminous. ' ' Borraginea. 



Flowers unsymmetrical. Stamens 5. Stigmas 5, Corolla 10-lobed. Ovary 1- 

 celled with 4^ erect basal ovules. Fruit baccate. Seed solitary, cotyledons pli- 

 cate in mucilaginous albumen. 



Flowers regular, infundibuliform, plicato-convolute in zestivation. Ovary 1-2-4- 

 celled, cells with 1-2 or rarely 4 ovules; placentae basalar. Fruit capsular or 

 baccate. Cotyledons conduplicately corrugated, albumen mucilaginous. Twin- 

 ing herbs or shrubs. 



Erycibea. 



Convolvulacea;. 



Group VII. Ovary 2-celled, placentae axile, carpels anterior and posterior, frnit baccate or 

 capsular. Seed albuminous. Herbs, shrubs or trees. 



Flowers regidar. Styles two. Fruit capsular, dehiscence loculicidal. Seed albu- 



minous. 



' HydroIeacea2. 



Flowers regular. Stamens generally isometrical with the lobes of the corolla. 



-Estivation often plicato-convolute. Fruit baccate, seed albuminous, often 

 compressed, lenticular. Solanacea?. 



Flowers irregular. Stamens 2 or 4 didynamous. Fruit usually capsular. Seed 

 albuminous. Scrofulariaceae. 



A cursory reTiew of these groups will shovr that although artificial in their mo^Q of con- 

 struction, thej yet bring together some very natural assemblages of orders. 



As regards the first, I almost doubt whether the orders composing it can be more 

 naturally placed, at least in the present state of the science. SalvadoracecB is, perhaps, an 

 exception, its affinities are still very obscure and do not appear to lie here, its present loca- 

 tion, therefore, is purely on account of its solitary, basalar ovule in which it coincides with 

 Plumhaginece. Plaiitaginem is another difficult order to place, but all Botanists agree in 

 keeping it near Plumhaginece, 



d^finit 



agrees 



The orders composing it are associated by nearly all 



Botanists. It is true Dn Arnott, in his temporary arrangement, has placed Sapotacem in his 



My 



EhQn 



in his linear series, nearly as I have placed them. The connexion between them and the associ- 

 ated orders of both Alliances I am unable to trace. With the exception of Ilic'mecB my series 

 corresponds with De Candolle's. 



The third, group though small, does not seem so natural. Oleacea; and Azimaceos are cer- 

 tamly near relations, not so JasminecE. Its true affinities must, I apprehend, be looked for in 

 the next, as suggested by Lindley, who places it in his Echial alliance among the orders forming- 

 my 4th group. Its right and left carpels, added to its peculiar placeutation, and mode of 

 development of the ovules, indicate a relationship to Verbenacees. 



The fourth group conveniently keeps together the 4 orders of which it consists. Though 

 all complex in their composition, they are so nearly allied and so interblended with each 

 other that nearly all Botanists seem to think it next io impossible to separate them without 

 doing violence to nature. Dr. Lindley, however, has seen reason to divide them, placing 

 GentianecPy Loganiacece and JpocynacecB in his Gentianal Alliance and JsclepiadecB in his 

 Solonal. I do not myself see the principles on which the separation is made, neither can I 

 trace the relationship between AsclepiadecB and SolanacecB. Gentianece clearly belongs to this 

 group, but to some extent possesses the ovarial character of the next, it therefore forms a 

 suitable transition. 



The fifth group is, with two exceptions, Lentihulariece and OrohanchaceeE, made up 

 of a series of orders which have always been recognized as closely allied. LentibularietB 

 which I hriner hf»ra nrlnflnaliw «n anfnnnt. nP ita ^'t(*a.rni]\arv ovarv and irreoTuIar flowers, exal- 



