988 CLASSIFICATION OF PLANTS. 



characters, 65 families of very unequal value. At tte commencement of his ' Fragmenta 

 Methodi Naturalis ' (1738) lie says: 'Diuet ego circa methodum naturalem inveniendam 

 laboravi, bene multa quae adderem obtinui, perficere non potui, continuaturus dum vixero ; 

 interea qua? novi proponam.' 



It was left to the sagacity of Bernard de Jussieu, when arranging the plants in the 

 gardens of the Trianon (in 1758), to seize upon the characters whereby the genera of Linnseus 

 should be methodized naturally under the primary divisions of Ray, and to the genius of 

 his nephew Antoine-Laurent de Jussieu so to characterize these subdivisions as that, with 

 but slight modifications, his ' Genera Plantarum secundum Ordines Naturales disposita,' 

 has, ever since its publication in 1774, retained its position as the basis of a complete 

 scientific classification of plants, and secured for its author the well-earned reputation of 

 the expositor of the Natural System. A.-L. de Jussieu's system comprised 15 classes, 100 

 Katural Orders, and about 1,740 genera. His classes are to some extent artificial, and 

 being founded upon modifications of the floral whorls alone, have been undervalued by 

 many botanists ; these modifications, though inconstant in each class, are excellent guides 

 to affinity, and (with the solitary exception to be noticed immediately) no better means 

 than these afibrd of subdividing either Monocotyledons or Dicotyledons into primary 

 groups, has hitherto been recognized. 



In 1763 Michel Adanson, who had studied in his youth under Bernard de Jussieu, 

 published his ' Families des Plantes,' a work of great originality and research, and which 

 is not only remarkable on this account, but as being the first complete system of Natural 

 Orders that ever appeared in print. These, however, though founded on natural characters, 

 being essentially artificial in construction, have had little influence in developing the 

 Natural System. 



In 1827 Robert Brown published his discovery of the direct action of the pollen-tube 

 on the nucleus of the ovule in two Dicotyledonous orders, Goniferce and Cycadeoe, thus 

 afibrding a sure basis for a subdivision of the class Dicotyledons into two primary groups, 

 the Gymnospermous and Angiospermous. Brown was also the first to propose the com- 

 bining of Jussieu's Orders into groups of a higher value, but subordinate to his classes ; 

 and in his various works he so improved both Jussieu's Orders and his Method, that 

 he ranks second only to that author and to Ray as the expositor of the Natural System of 

 Plants. 



In 1813 Augustin Pyrame de Candolle, then Professor of Botany at Montpellier, pub- 

 lished the first edition of his ' Theorie Elementaire,' in which he proposed a modification of 

 Jussieu's classes, and of his sequence of the Orders, which is in some respects an improve- 

 ment, and in a few the reverse ; as where he breaks up the Acotyledons, and places the 

 vascular Orders of these amongst Phssnogams, calling them Cryptogamous Phsenogams. A 

 great improvement was his combining the Diclinous and Apetalous Classes of Jussieu. The 

 first edition of the ' Theorie ' contains 145 Orders ; the second, which appeared in 1819, 

 contains 161 ; and the last, edited by his son, A. de Candolle, in 1844, and which represents 

 his latest views, contains 213 Orders : in ii the vascular Cryptogams are removed from the 

 Pheenogams, but the Gymnospermous division of Dicotyledons is not recognized. 



The sequence of Orders given in this work is for the most part that proposed by De 

 Candolle, and is that . followed in the universities and schools of England and its depend- 

 encies, in America, and throughout a great part of the continent of Europe. • The arrange- 

 ment of De Candolle, given at p. 165 of this work, is that of the first edition of the ' Theorie 

 Elementaire.' The following is his matured plan, as given by his son : — 



