128 Mr. J. Miers on the Styracese, 



Prof. Decaisne considered the Styracece nearly allied to Alan- 

 ffiacets ; but in this conclusion there can be little doubt that he 

 held SymplocacecB in view, that being nearly the position I have 

 assigned to the latter family. 



Prof. Miquel (in 1856), in Martius's 'Flora Brasiliensis/ 

 adopted the order Symplocaceaj which he there described as 

 distinct from Styracece , according to my suggestion, but without 

 offering any comment on the matter. 



Finally, Prof. Agardh, in a work just published (1858), gives 

 quite a novel view of the affinity of the Styracece^ which he is 

 disposed to consider as a gamopetalous form of the Tiliacecey and 

 more especially of the Elceocarpece. He seems to accord in the 

 separation I have proposed* of the Styracece from the Symploca- 

 cea, and suggests the probability that the latter family may prove 

 to be a gamopetalous form of his group Carpodetece, which he 

 arranges near Cornacece. 



From this history it will be seen how constantly and how 

 widely the conclusions of botanists have varied in regard to the 

 affinities of these two small groups of plants ; and this uncer- 

 tainty shows a necessity for fixing their characters upon some 

 more definite basis than has heretofore been attempted. 



The affinities of the Styracece, and the mutual relationship of 

 the several genera there associated, have hitherto been founded 

 on the more trivial characters of the relative number of the sta- 

 mens, and their cohesion at their base upon the petals, which at 

 this point often become pseudo-gamopetalous by mere aggluti- 

 nation ; while, on the other hand, no stress has been laid on its 

 other far more essential carpological characters. Although it be 

 true that in forming our judgment regarding affinities, we should 

 trust to the assimilation of several features rather than to few, 

 still the precept of Jussieu should be attended to, that the most 

 constant elements are to be observed in the principal organs of 

 reproduction, especially those of the ovary and seed, and that 

 these characters therefore should hold a prior claim over other 

 concomitant, more variable, and less important features, in our 

 investigations into the mutual relations of plants. This rule 

 appears to have been wholly disregarded, in respect to the Sty- 

 racece, by nearly all the botanists who have written on this 

 family. Notwithstanding that the diagnostic features of the 

 genera are severally drawn up in a very elaborate and careful 

 manner by Prof. A.DeCandolle, in his excellent monograph of the 

 order, the difi'erential characters of the tribe Symplocece (Prodr. 

 viii. 246) and those of the Styracece {p. 259) are confined wholly to 

 the dissimilarity in the aestivation of the corolla, the number of 

 stamens, and the size of the cotyledons in relation to the radicle : 

 no comparison is made of the structure of the ovarium, which is 



