as distinguished from the Symplocacese. 143 



in Striffiliaj though they do not appear to have attracted his 

 attention. 



It is difficult to conceive the grounds upon which Prof. Agardh 

 regards the Styracea (separated from SymplocacecR) as being 

 more immediately allied to the ElmocarpecBj among Tiliacece^. 

 By his novel system, the affinities of different families of plants 

 are best demonstrated by the form and mode of development of 

 the ovules. As this character is liable to be modified by many 

 circumstances, it can never retain the importance there attached 

 to it, because if it be employed as a primary mark of distinction, 

 it will often lead to error ; but it is nevertheless of considerable 

 value as an accessory feature, which has not hitherto been suffi- 

 ciently attended to : if, however, we combine with this its essential 

 concomitant, the position and distribution of the placentary por- 

 tions of the component carpels of the ovary, the affinity in ques- 

 tion ceases to be apparent. In his group of the Elceocarpea, in 

 which he includes the Tricuspidariece of Endlicher, the union of 

 the component carpels constituting the ovary is complete, and 

 their placentary margins all unite in a solid axis in the centre, 

 the ovules being attached in collateral pairs, in each cell, upon 

 the middle of this axile column. No similar structure exists in 

 Styracece. In the Styracinea, as I have explained, the numerous 

 ovules, generally in three series, are seated on a central abbreviated 

 column, which has no direct connexion with the style, with the 

 raphe ventral in the erect ovules, superior in the horizontal ones, 

 and dorsal in the lower row, as Prof. Agardh admits : but this 

 seemingly deviating position of the raphe is merely the effect of 

 their resupination on their funicles, produced by mutual pressure 

 during growth ; for if each ovule be separately brought into one 

 similar angle of radiation from the axis, the raphe in all of them 

 will be seen in the same facial direction. As an example of the 

 different mode of development in Elcsocarpea, I will cite what I 

 have observed in Aristotelia : the ovules, two in each cell, are there 

 attached collaterally a little below the summit of the axis ; they 

 are naturally at first cupuliform, as Prof. Agardh shows in tab. 21, 

 fig. 7; and in the progress of their growth they probably become, 

 as he says, mutually heterotropal, — that is, one growing upwards, 

 the other downwards, with the raphe towards the axis in both 

 cases. In confirmation of this, I have observed that, at the 

 period of expansion of the flower, the ovules, by the effect 

 of pressure against the cavity of the cell, become twisted round 

 upon their funicles, so as to appear superposed, — the left ovule 

 becoming superior, with its singularly curved chalazal point 

 directed to the dexter side, the right ovule becoming inferior, 

 with its chalazal point turned to the sinister side j and this relative 

 * Theor. Syst. Plant, p. 269. 



