334 MORPHOLOGY. 



a monstrosity, fin aberration from the normal course of development ; 

 and there exists no principle here, by which to judge how far the plant 

 deviates from its type to produce new forms, or how far the usual form 

 is to be carried back to simpler elements. Assuming that the coronet 

 of Narcissus consists of independent foliar organs blended with the 

 leaves of the perianth, may not they be multiplied in the doubling of 

 the flower as well as the others, and, since their adherence to the peri- 

 anthial leaves is merely a characteristic of the special nature of the 

 flower, be isolated from each other and become blended one to each peri- 

 anthial leaf? Monstrosities prove nothing at all here, but merely render 

 probable ; the only perfectly sure decision is to be obtained here, as every- 

 where, by tracing the development. In the paragraphs I have connected all 

 these appendages to the analogy with the ligule, which is certainly war- 

 ranted by the development for some forms, e. g. Narcissus, Silene, &c. 

 In others, indeed, no such analogy exists, as in the Passifloracea ; in very 

 many forms we have no accurate investigations whatever, as in Par- 

 nassia and in the Stapelice. The fleshy portions of the latter form the 

 transition to the thick fleshy papillce, such as occur, for instance, in 

 such strange forms in the labellum of the Orchidacece ( Oncidium) ; on the 

 other hand, the filiform appendages of Passiflora, which indeed belong 

 partly to the disc, are more nearly allied to those tufts of hair occurring 

 in more definite places, and with more definite colour and form, which 

 are called beards (barba), for instance, on three of the perianthial leaves 

 of Iris. 



Finally, I will observe, that I have referred to the terms employed to 

 describe the total form of the floral envelopes, as of general application, 

 although most of them are only mentioned in isolated special cases, with- 

 out involving any thing at all specific for one or other kind of floral 

 envelope. Most of the words are very readily understood, some more 

 difficult ; e. g. hypocrateriforme, which will not be easily understood by 

 any one who has not seen in old collections or old paintings the form of 

 a flat plate supported on a long stem, on which wine-glasses were placed 

 in the middle ages ; floral envelopes are called rotate when they expand 

 the separate organs equally or almost equally from the point of attach- 

 ment in one plane. It also appears to me to be a mistake when a part of 

 these expressions are applied solely to floral envelopes when the leaves 

 are coherent, which merely gives rise to the necessity of imagining a 

 new word for the condition with free leaves. What it is here required to 

 name is the general contour, and therefore it is no more necessary to 

 regard the subordinate divisions in these, than in the subdivided flat 

 organs, such as the lobed and pinnatifid leaves. In rotate, salver- 

 shaped, &c. corollas, however, the limb is always divided, which never 

 is the case in a wheel or salver ; and in this mode of naming we cannot 

 convey ideas of degrees of more or less, so that in the corolla of Lychnis 

 and Dianthus the words corolla (pentapetala) hypocrateriformis an- 

 fivver the purpose very well. 



149. Five kinds of floral envelopes are distinguished. When 

 all the foliar organs are similarly or nearly similarly developed 

 in a circle of one evident form, colour, and structure, they are 

 described under the general name of perianth, the single organs of 

 which are called perianthial leaves. If in the floral envelopes of 

 one flower we can distinguish two circles differing in form, colour, 



