102 MR. G. BETsTHAM OJS' MTRTACE-aE. 



more apparent ones, those the most useful for distinguishing great 

 groups in one of the three great centres of the order, South Ame- 

 rica, tropical Asia, and Australia, will not always hold good in 

 the others. The consequence is that no general arrangement 

 can be satisfactory, if founded on the species of one only of those 

 regions ; and of this description are all those that have been pro- 

 posed since the ' Prodromus ' of DeCandolle. His distribution 

 was very good, considering the scanty materials he then had at his 

 disposal, but has required very great modifications after the large 

 accessions we have since accumulated. This rearrangement has 

 been more or less worked out for each of the three regions in two 

 opposite directions. An enormous multiplication of genera and 

 species has been proposed by O. Berg for the South- American 

 Myrtacea?, by Blume for those of tropical Asia, by J. C. Schauer 

 for the Australian ones, and by Brongniart and Grris for the 

 smaller group of IN'ew Caledonia ; whilst reconsolidation has been 

 more or less effected by Grisebach for tropical America, by Wight 

 and A. Gray for tropical Asia, and by F. Mueller for Austraha; 

 and, whether disruption or consolidation have been the guiding 

 principle, the characters made use of in such widely spread groups 

 as Myrtus^ Eugenia^ and their allies have often been different, ac- 

 cording to whether the American, the Asiatic, or the Australian 

 species have been chiefly had in view. AH these difficulties which 

 the systematist has to encounter are not a little enhanced by the 

 tedium of examining ovaries and seeds in an Order where their 

 resinous or woody nature often retjuires much boiling to soften 

 them, and where especially it is so unsafe to conclude upon the 

 structure of one species from the examination of an apparently 

 closely allied one — where in such vast natural and almost uniform 



Eucalyptus^ and Mel 



anv 



have a totally exceptional anther, or ovary, or embryo. 



Having myself on former occasions examined a large number 

 of American and Asiatic species, and having now been obliged 

 carefully to go through the whole of the Australian ones, as 

 well as many species of the tropical genera which had not pre- 

 viously come under my eye, I have thought it might be wortu 

 prefacing my notes on the limits of individual genera by a fe^ 

 remarks on the stability and instability of the several characters 

 observed, w^hich I shall take in the usual systematic order, com- 

 mencing with the flower and fruit and then passing to the organs 

 of vegetation. 



