682 SCROPHULARIACE^. [Perigynous Exogens. 



puUnnate mass in the fork. Fruit capsular, seldom berried, dicarpellary, 2- celled 

 sometimes with 2 entu'e or bifid valves, sometimes w-ith 4 entire ones, sometimes open- 

 inc by pores or lids, very rarely almost indehiscent ; dissepiment parallel or opposite to 

 the valves, finally loose in the centre, or altogether. Placentae adhering to the dissepi- 

 ment, sometimes when matm'e separate and forming 1-2 central columns. Seeds 

 indefinite, rarely definite, albuminous ; embryo orthotropal, heterotropal, or antitropal, 

 but slightly cm'ved. — Bcntham. 



The capsular monopetalous genera of Dicotyledons, ^vith a superior ovary, albmninous 

 seeds, and irregular diandrous or didynamous stamens, were separated by Jussieu into 

 two Orders, which he called Scrophulariae and Pediculares, distinguished from each other 

 by the dehiscence of their fruit : the former being septicidal, and the latter loculicidal. 

 Bro\vn, in his Prodromics, pointed out the insufficiency of this character, which is often 

 not even of generic value, and he combined the Orders of Jussieu under the common 

 name of Scrophularineae (Figworts). This opinion has been adopted by subsequent 

 WTiters, with the exception of De Candolle, who, in Duby's Botanicon Gallicon (1828) 

 adheres to the old division of Jussieu, the names being changed mto Antirrhinese and 

 Rhinanthaceee. Notwithstandmg the almost universal assent to the identity of the 

 two Orders of Jussieu, some separations have been made upon different principles from 

 those of that learned Botanist. Thus Broomrapes have been distinguished by himself ; 

 Gesnerworts by Nees Von Esenbeck ; and Melampj-racese by Richard. The two former 

 are adopted by all Botanists ; the latter group has not been generally received. 

 I formerly admitted it, upon the gi'ovmd of its definite ascending seeds and inverted 

 embryo ; but subsequent observation led me to think that by excludmg from the 

 character all consideration of the number and du-ection of the seeds, an Order would be 

 formed, agreeing in a peculiar habit, and in the radicle of the embryo not being 

 presented to the hilum, to which the name of Rhinanths might conveniently be retained. 

 According to this \'iew of the subject, Figwoi-ts would include no genus the embryo 

 of which is not orthotropal, and m Rhinanths it could be antitropal or heterotropal. 

 But although the attachment of the seeds of Rhinanths is generally lateral, yet some- 

 times the radicle points to the hilum ; but it is more generally removed from it. The 

 ovules are never fewer than 2 ui each cell, often numerous, and there are sometimes, 

 though rarely, 2 o^^lles only in the ovary of some of the tribes of Figworts. And 

 therefore the ground for separating Rhinanths from Figworts sinks from under us. 



The number of sjiionjTnous names above quoted, shows into how many more 

 supposed Orders the old Scrophularise have been broken by one author or another. 

 The whole matter has, however, been investigated by Mr. Bentham, who has treated 

 the question in both a philosophical and practical way, and who concludes that in fact all 

 the supposed Orders are really sections of one great Natm'al Order. 



Mr. Bentham remarks that the nearest Order to Figworts is undoubtedly that of 

 Nightshades, through the medium of Salpiglossids ; so that it becomes necessary to 

 separate them by a purely artificial distmction, considering as Nightshades such genera 

 as have a plaited corolla and 5 stamens, and as Figworts all those in wliich either the 

 fifth stamen is wanting, or the aestivation of the corolla imbricated. The line would 

 thus be drawn between Petunia and Salpiglossis, two genera closely allied in habit. In 

 the first, however, the decidedly phcate corolla and 5 stamens show it to be a true 

 Nightshade, whilst the shght irregularity of the coroUa and the declinate very imequal 

 stamens, indicate an approach to Salpiglossis, which, being always didynamous, with an 

 imbricately eestivating, or obscurely plicate corolla, is a genuine Figwort. Among 

 Verbascese the genus Verbascum which is pentandrous, and Celsia, because it cannot be 

 separated from Vex'bascum, have usually been referred to Nightshades, although no 

 plants nearly alhed to Verbascum occur in the latter Order ; but the aestivation of the 

 corolla, besides the general habit, leave no doubt that Bartling and others are right in 

 classing these genera among Figworts. A better reason seems to me to be furnished 

 by the manifest tendency to lose a part of the stamens, which occm's in Verbascum. 



From the other Orders of this Alliance the Figworts are sufficiently well distin- 

 guished. They differ from Pedaliads, Gesnerworts, and Crescentiads in their placenta 

 never being parietal ; from Bignoniads and Acanthads in their albuminous seeds and 

 small cotyledons ; from Butterworts in then* axile, not free central placentae. 



Some Figworts approach Broomi'apes in the peculiar habit of that Order, especially 

 the Buchnereae, among which most, if not all, the Strigas are parasitical, and Buchnera 

 hydrabadensis is actually leafless hke a Broomrape. 



The two tribes of Mitrasacmeae and Buddleae approach Loganiads in their leaves 

 being connected by a transverse line, which occasionally expands in the form of stipules; 

 but they differ in their flowers being hu'egular, at least in aestivation, one lateral 

 lobe being outermost, whilst the upper one is innermost. In all Loganiads which 1 

 have examined, the aestivation is either regularly convolute or valvate. The irre- 



