264 THE JUSSIEUS AND THE NATURAL METHOD. 



After duly considering their relative value, be recognized the germination of the 

 seed and the respective arrangement of the sexual organs as the two priiicipa] 

 and mof^t invariable ; he adopted them, and made them the basis of the arrange- 

 ment which he established at Trianon in 1759." 



Thus the solution of the problem I had proposed is seen to have little diffi- 

 culty ; for Laurent himself tells us, as well in these notes as in the preface to 

 his Genera Plantarum, that it is to Bernard we owe the discovery of the prin- 

 ciple of subordination of characters. "This inequality of characters had not 

 escaped the excellent author of the Orders of Trianon, neither the subordinate 

 part of the more variable, nor the importance of the more constant, nor the dig- 

 nity of the embryo and the sexual organs, nor the affinity of the genera and 

 orders which are associated with one another by these primary indications. The 

 families which he has established are, in general, strictly natural, and confoima- 

 ble to these principles." At a still later period, he styles the Catalogue oj 

 Trianon, that mature result of the long meditations of Bernard, "the most solid 

 monument of his renown." Nor does M. Adrien de Jussieu, though disposed 

 by a natural bias to incline the balance rather to the side of his father than his 

 uncle, bear a different testimony : "I have beneath my eyes the manuscript cata- 

 logues of Bernard : there are two of them ; that which was printed at the head 

 of the Genera, and another still longer, in which are enumerated, in connection 

 with the name of each kind, the species according to the Linngean nomenclature, 

 with a brief synonomy of former authors. But the whole is limited to a series 

 of names, without a word of development or explanation. Such as they are, 

 however, they evince that Bernard de Jussieu had established the principle of 

 the subordination of characters, and had determined those to which must be 

 assigned the first rank ; an immense step in advance, and sufficient in itself to 

 immortalize him who conceived it." * * * * "But does this embrace," asks 

 M. Adrien with reason, "all that we find in the Genera Plantarum V In reply, 

 let us briefly examine that work. At the time of its appearance, botany possessed 

 20,000 plants, of which more than half had been unknown to Bernard — those 

 of Commerson, of Dombey, of Forster, of Forskal. The author distributes these 

 20,000 plants into a hundred orders; these hundred orders into 1,754 genera;* 

 to each of these orders and genera are assigned its characters, and to all these 

 characters their due valuation and weight. 



The author divides the characters into three classes : The first class, essential, 

 constant, uniform in all the orders, and drawn from the most important organs, 

 the number of lobes or cotyledons of the embryo, the insertion .of the stamens 

 or their arrangement in relation to the pistil, the situation of the staminiferous 

 corolla; the second class, general, nearly uniform in all the orders, or only vaiy- 

 ing by exception, and drawn from organs less important — the presence or defect, 

 whether of the calyx or of the non-staminiferous corolla, the structure of the 

 corolla considered as monypetalous or polypetalous, the relative situation of the 

 calyx and the pistil, finally the presence or absence of the perisperm ; the third 

 class, sometimes uniform and sometimes variable, now furnished by one organ 

 and now by another, the calyx monophyllous or polyphyllous, the ovary simple 

 or multiple, the number, proportion, connection of the stamens, the number of 

 cells of the fruit and its manner of opening, the position of the leaves and flowers, 

 &c., &c. By virtue of this classification of the signs, Laurent has always before 

 him the principle which controls the arrangement of plants. It only remains 

 to respect everywhere this first classification, which gives the other. Let no 

 character of a genus intrude into the definition of an order, nor of an order into 

 the definition of a genus. The least inversion produces dissonance in the natural 

 order. By this system the method is seen, more clearly than ever before, to be 

 ih^ science of characters. There are found to be laws by which these characters 



* Add 150 genera which are supernumerary, or of doubtful place (planta inceitcE sedis.) 



