178 AN ACCOUNT OF THE GENUS AEGEMONE. 



support that estimate; the "species" of this paper are in reality 

 aggregations of forms that probably most botanists would recognise 

 as specifically distinct. What canons are applied in limiting the 

 species admitted in the Genera Plantarum or in the Histuire des 

 Plantes we have no means of judging; it is, however, evident, 

 from the citation there of A. Hunnemannii as distinct, that in 

 the Natuiiichen PjianzenfamiUen the limitation is less rigid than 

 that employed here. 



As regards the claim to recognition oi A.fniticosa, dispute, as 

 has been already said, is impossible. Its fruticose habit, its holly- 

 like leaves, its capsules dehiscing nearly to the base, separate it 

 unmistakably from all the others. In the rest we find the habit 

 herbaceous, the leaves thistle-like, the capsules dehiscing only in 

 the upper part. Indeed, it is a matter to be thankful for that 

 generic rank has not as yet been claimed for A.fniticosa, seeing 

 that the character afforded by its fruit is exactly that on which 

 alone depends the separation of Pioemeria from Papaver § PJueades, 

 and of Cathcartia from Meconojjsis. 



When, however, we examine the remaining forms, considerable 

 difficulty is experienced. The general characters derived from 

 habit and foliage are somewhat variable within each, and certain 

 of the forms simulate others in a remarkable manner. Thus A. 

 stenopetala repeats the habit and foliage usually present in A, mexi- 

 cana, while A. inteniiedia in its southern form repeats those present 

 in A. ochroleuca, and in its northern form (characteristic of the 

 western prairies) those of A. plattjceras. A. alba usually resembles 

 A. ochroleuca, but sometimes imitates the northern form of A. inter- 

 viedia, while A. rosea (the Chilian plant) imitates now A. hifspida of 

 the Eocky Mountains, and now A. (jlnuca of the Sandwich Islands. 

 If these general characters are to be relied on, there is nothing to 

 be said against the view that would reduce the number of species 

 of Argemone to two, viz., A. fruticusa and a second composite and 

 very variable species. If, indeed, tlie commonly received reduction 

 of A. alba to A. mexicana be insisted on, this wider reduction 

 becomes a logical necessity. When, however, the subject is more 

 closely investigated, we find that the general similarity in this last 

 instance is rarely very great, and is never so marked as in the case 

 of some of the forms usually accepted as distinct ; there is, more- 

 over, always a characteristic difference in the disposition of the 

 bracts, the shape of the sepals, the size, shape, and colour of the 

 petals, and, above all, in the shape and consistence of the fruit. 

 Indeed, beyond the fact that both are Argemones, there is nothing 

 in favour of their reduction to one species. The same remark 

 applies with equal force to the identification of ^. intermedia with 

 A. ochroleuca. 



(To be contiuueJ.) 



