RULES FOR INVESTIGATION. 



9 



two first subdivisions, it must necessarily be one of those in 

 the third subdivision. In this there are only three genera, 

 viz. Phaca, Biserbula, and Astragalus. The plant in question 

 not having a half-two-celled inflated legume, it cannot be the 

 Phaca ; it must therefore be either the Biserbula or Astragalus. 

 The gibbous shape of the legume, leaves no doubt of its being 

 an Astragalus ; but as the species of this genus are extremely 

 numerous, in order to determine the species, we must pay 

 very particular attention to the minute circumstances by 

 which they are distinguished. The species of Astragalus are 

 divided into not fewer than seven subdivisions. The speci- 

 men before us cannot belong to the six .first subdivisions, 

 because the first consists of Astragali, with leafy erect stems 

 and axillary flowers ; the second of leafy erect stems, with 

 spikes cylindrical, axillary, sessile, or nearly so ; the third of 

 leafy erect stems, with spikes and racemes peduncled ; the 

 fourth of leafy diffuse stems ; the fifth of leaflets placed in 

 whorls ; and the sixth of a naked scape, without a leafy stem. 

 Our plant must therefore be a species of Astragalus, in the 

 seventh subdivision, containing shrubby plants with the 

 petioles permanent, and becoming spinous. As the species 

 of Astragali in this subdivision are numerous, we must be 

 very careful in examining all the specific characteristics. The 

 leaflets being in ten pairs, determine it to be either the Astra- 

 galus Longifolius, or Astragalus Pugniformis. It cannot be 

 the latter, which has oblong mucronate glabrous leaflets ; 

 therefore there is not the least doubt that the plant in question 

 is the Astragalus Longifolius, from the lanceolate mucronate 

 glabrous leaflets, the flowers being in oblong heads, and the 

 calix having filiform villous teeth. 



STARWORT. Aster. 



Upon the first sight of this plant, we refer it to the class 

 Syngenesia, because of its composite flowers. As this class 

 is subdivided into five orders, we soon perceive, by a careful 

 inspection of the florets, that it cannot belong to the first 

 order, which consists of plants having florets all hermaphrodite 

 and fertile. Neither can it belong to the third order, which 

 consists of plants with the florets of the centre hermaphrodite, 

 and those of the ray neuter and barren ; neither can it be one 

 of the fourth order, which consists of plants having the florets 

 of the centre hermaphrodite and barren, and those of the 

 margin, female and fertile : nor can it belong to the last order, 

 which consists of plants with the florets separated by a proper 

 perianth, containing one or more florets, and placed within 

 a common calix. The florets of the disk, in the specimen 

 before us, being hermaphrodite, and those of the margin 

 female, and all fertile, leave not the least shadow of doubt 

 of its being in the order Polygamia Superflua. This order 

 has three grand subdivisions. In the first, all the florets are 

 tubular, which is not the case with the plant in question; 

 in the second, the florets are semi-floscular, and somewhat 

 two-lipped, circumstances not in our specimen. The radiate 

 florets of our specimen, is the striking characteristic of the 

 third subdivision, and is decisive of its being in this sub- 

 livision. Its naked receptacle is a proof that it cannot be 

 Boltonia, Siegesbeckia, Anthemis, Achillea, Tetragono- 



Jca, Ximenesia, Georgina, Relhania, Pascalia, Buphthal- 



mum > Rhanterium, Amellus, Tridax, Rosenia, Verbesina, 



.cnlechtendalia, Galinosogea, Zinnia, Balbisia, nor Starkea. 



s simple down, in union with its naked receptacle, confines 



to the Doronicum Arnica, Inula, Erigeron, Solidago, Cine- 

 ria, fcenecio, Tussilago, Aster,' and Boebera. Its imbricate 



lix, with the lower scales spreading, and the florets of the 

 margin more than ten, restrict it entirely to the Aster. The 



VOL. I. 1. 



species of this genus are very numerous, and are divided 

 into seven grand subdivisions. It cannot belong to the first, 

 which consists of species more or less shrubby ; nor to the 

 second, consisting of herbaceous plants, with the stem one 

 or two flowered ; nor to the third, containing herbaceous 

 species, with linear or lanceolate leaves very entire ; nor to 

 the fourth, consisting of herbaceous plants, with heart-shaped, 

 ovate, and serrate leaves ; nor to the fifth, containing herba- 

 ceous plants, with ovate very entire leaves ; nor to the sixth, 

 which contains herbaceous plants, with lanceolate leaves, the 

 lower ones more or less serrate. Our plant being herbaceous, 

 with pinnate leaves, must belong to the last subdivision, in 

 which there are only two species, the Aster Aurantms, and 

 Aster Pinnatus. The pinnate leaves, with the midriff 

 winged, leave no doubt of 'its being the Aster Pinnatus; 

 and the additional circumstances of the leaflets being linear 

 wedge-form, and serrate at top, put the matter beyond all 

 controversy. 



TRUMPET-FLOWER. Bignoma. 



The stamina of this plant being two longer and two 

 shorter, is sufficient to determine it to be in the'Didynamia 

 class ; and the seeds being inclosed in a capsule, shew it to 

 be of the Angiospermia order. The plants of this order are 

 numerous, and subdivided into seven parts, all of which, 

 except the last, depend upon the form of the calix. The 

 plant in question is not in the first subdivision, which consists 

 of plants with undivided calixes ; nor in the second, consist- 

 ing of bifid calixes ; nor in the third, containing plants with 

 three-cleft calixes ; nor in the fourth, which contains plants 

 with four-cleft calixes ; nor in the sixth, containing plants 

 with many-clefted calixes ; nor in the seventh, containing 

 plants with more than one petal. Our plant having a five- 

 clefted calix, there can be no doubt of its belonging to the 

 fifth subdivision. The genera in this subdivision are numer- 

 ous, which are generally distinguished from each other by the 

 capsule and corolla. Our specimen has a two-celled capsule 

 and a Campanulate Corolla ; the union of which particulars, 

 restrict it to the following genera : Capraria, Digitales, Big- 

 nonia, Incarvillea, Ruellia, Sesamum, and Gloxinia. The 

 imbricate-winged seeds and the filaments, with the rudiment 

 of a fifth, shew it can be no other than the Bignonia. The 

 species of this genus are very numerous, and are subdivided 

 into seven parts ; all of which depend upon the shape or 

 number of the leaves. The plant in question consequently 

 cannot be in the first subdivision, which contains plants with 

 simple leaves ; nor in the second, consisting of plants with 

 two leaves on each petiole ; nor in the third, having plants 

 with three leaves, on a common stalk ; nor in the fourth, with 

 leaves in finger-like divisions; nor in the fifth, with pinnate 

 leaves ; nor in the seventh, with doubly pinnate leaves. The 

 decompound leaves of our specimen, shew it to be in the sixth 

 subdivision. In this there are only three species. The de- 

 compositely pinnate leaves, with knotty petioles and axillary 

 corymbs, shew it to be no other than the Bignonia Alba. 

 The additional circumstance of its having white flowers on a 

 many-flower-forked axillary peduncle, puts the question be- 

 yond all dispute. 



SEDGE-GRASS. Carex. 



Upon a minute investigation of the apetalous flowers of 

 this plant, we discover that some are male and others female, 

 growing from the same root ; but we are not able to find any 

 hermaphrodites; consequently it must belong to theMonoecia 



