58 



ILLUSTRATIOxNS OF 



I 



Salpiglossis. 



Upon a former occasion [huj. op. i. p. 172) many reasons were 

 adduced to show why the tribe of the'Salpiglossidece, as constituted 

 by Mr. Bentham (DC. Prodr. x. 190), could not be maintained, 

 and I proposed to limit that tribe simply to Salpiglossis , Browallia, 

 Leptoglossis, and a new genus Pteroghssis, all being distinguished 

 by then- singularly dilated stigma and the peculiar mode of esti- 

 vation of the corolla. A careful examination of Leptoglossis 

 schwenhoides has since then offered reasons for placing that 

 genus among the Petuniecs. The Salpiglossidece, however, as 

 thus limited, are evidently most intimately allied to the Petuniecp, 

 agi-eemg with them in a somewhat similar form of stigma, the 

 development of their stamens, their capsular fruit, and the very 

 spiral form of the embryo in Salpiglossis, and differing from them 

 only m their didynamous stamens and the estivation of the 

 corolla. The didynamous arrangement of the stamens does not 

 appear to me to offer a sufficient reason for keeping them in an" 

 ordinal point of view apart from the Petunieee, and for retaining 

 them m the Scrophuhriacea ■ indeed in the Petuniea and Nico- 

 tianece, we find an evident tendency towards a didynamous struc- 

 ture, for one of the stamens is constantly shorter than the others 

 which are m two pairs, whHe the anther of the fifth is always 

 somewhat smaller, and frequently almost sterile; and on the 

 other hand, I have observed occasionally in Salpiglossis a fifth 

 fertile stamen, showing a disposition to return to its normal con- 

 dition ; and I have also before me an instance of a flower with 

 three pairs of stamens, varying in length, with a seventh shorter 

 one, the anther of which, though smaller than the others, is fer- 

 tile. The position of the Salpiglossidea in the natural system 

 appears to me therefore manifestly in the family which I propose 

 to c2i\]Atropace<s, or if considered only as a suborder, Atropinetje 

 according to the arrangement there shown [lac. ciL p. 165) 



There is httle in the genus Salpiglossis that calls for observa- 

 tion ; one peculiar feature however claims attention, the singular 

 form of its pollen-grains : these are comparatively large and rea- 

 dily distmgmshed under a common lens, each granule consisting? 

 of four agglutinated spherical globules similar in form to the 

 simple pollen-grains of most Solanacece and ScrophulariacecB : 

 three of these globules are on the same plane, the other beinff 

 superimposed in the centre, thus forming a sort of rounded tetra- 

 hedron and they adhere so completely that they cannot be sepa- 

 rated withou bursting. The fact is noticed by Mr. Hassall in 

 his memoir "On the Structure of Pollen- (Ann. Nat. Hist. viii. 



y- 



curious a circumstance is not singular 



Leschenaultia in Gt 



