Mr. J. Miers on the Calyceracese. 183 



consequently been regarded as holding an intermediate position 

 between that family and the Dipsacece, with which it also accords 

 in its capitate involucrated inflorescence, its monopetalous co- 

 rolla, its inferior 1- celled ovary with an adnate calyx, its undi- 

 vided style and simple stigma, and its solitary suspended ovule, 

 — its fruit being likewise a monospermous achsenium, crowned 

 by the persistent teeth of the calyx, and the embryo of its seed 

 imbedded in an ample albumen. It differs, however, from the 

 Dipsacece in the peculiar venation of the corolla, in having its 

 filaments united at the base into a monadelphous ring, and in 

 its combined half-syngenesious anthers. 



Mr. B. Clarke * first proposed to separate these families, by 

 some distance, in a natural system founded principally on the 

 normal position of the carpels and the relation of the raphe to 

 the placenta. Under his arrangement, in the monopetalous 

 proterocarpous division of Exogens, he places the Calyceracece in 

 his Tetragonal alliance with Valerianacece and Dipsacece, while 

 Composite appear in his Myrtal alliance, the Onagrarial alliance 

 being interposed between them. It is needless to say that the 

 grounds of arrangement in this system are most feeble, as the 

 more important considerations of floral and carpological struc- 

 ture, as well as general habit, are completely placed out of view. 

 As no other reason is adduced to justify the separation of the 

 two families in question, the validity .of their close affinity still 

 remains unimpeached. 



Prof. Agardh has likewise proposed the separation of the Caly- 

 ceracece from the Composite, under a still more singular system 

 of arrangement, also founded principally upon the mode of de- 

 velopment of the ovules in different plants. He places Compo- 

 site at the end of an alliance which embraces Cycadacece, all the 

 Coniferce, Ephedracece, Proteacece, and Bruniacece, and places 

 Calyceracece in another alliance of the most heterogeneous kind, 

 comprising Plantaginacece, Primulacece, Dipsacece, and Gunnera- 

 cece, the latter family interposing between Calyceracece and Com- 

 posite. The principal reason given for this approximation of 

 Gunneracece with Calyceracece is the fact of the single ovule 

 being in both cases suspended from the very apex of the cell, 

 while in Dipsacece the point of its suspension is somewhat lateral. 

 The view of natural affinities expounded in his workf will not 

 stand the test of investigation ; when applied to the case under 

 consideration, we see that Gunnera, with its monochlamydeous, 

 often dioecious flowers, which are spicate or racemose and ebrac- 

 teated, its two simple stamens, its divided style with plumose 

 stigmata, its drupaceous fruit and very minute embryo in copious 

 albumen, offer a combination of characters that place it at a great 



* Ann. Nat. Hist. 2 ser. xi. 454-456. f Theoria Syst. Plant. 



