238 PIULOSOPHICAL NOTES ON 



two stamens becoming fused into one, either purtially, as might be 

 those of the Hornbeam, or completely, as might be those of 

 Hepatica, Heartsease, and others. 



In the "N'uovo Giornale Botanico " for April 1888, Tab. 15, 

 Fig. 7, are shown two stamens of Ficus fused by their filaments, 

 lea^dng the anthers separate, as in the Hornbeam. 



In Tab. 14, Fig. 12, the same occurs in Verbascum. 



In Vol. 18, Tab. 3, of the same journal, a number of terato- 

 logical anthers are given from Lonicera, which support the notion 

 of the anther originating in an incurving of the margins of the 

 leaf or leaflet, as in the sori of Cryptogr amine crispa and others. 

 The whole fertile frond of this fern is like a series of anthers. 



The possible origin of the two-celled anther out of the union 

 of two one-celled anthers is also seen in the divergent anther cells 

 of Penstemon puhescens, of several Labiatfe, of Monarda, 

 Calamintha, of Salvia, &c.* 



It would seem easy for the approximation of such one-celled 

 anthers, and subsequent connation, to give origin to those of 

 JPyrola, Isopyrum, Liriodoidron, Asarum, and others. "f 



In Schizandra coccinea, the anthers are one-celled on the 

 opposite margins of petaloid filaments. The contraction of such 

 a filament would naturally approximate the one-celled anthers, and 

 make a two-celled anther. 



The anther cells having become connate, to form a two-celled 

 anther, may afterwards group themselves in various ways, by 

 fusion of their filaments, as we see them in Candollea cunei- 

 formis. 



The circular anther of Cyclanthera is evidently a fusion of 

 many anthers forming a ring, with one continuous circular 

 dehiscence. 



In the Cucurbitaceae there is often a combination of two-celled 

 and one-celled anthers, each twisted into the shape of an S for 

 want of space. 



In all anthers, whether one-celled or two-celled, the pollen 

 grains might be comparable to the globular hairs of the Cheno- 

 podium petiolare leaf, and to the ovules of Nymphcea, Butomms, 

 and others, in which the whole carpellary surface is ovuliferous. 



* See Asa Gray's " Struct. Bot," pp. 255 and 256. - 

 t See p. 252 of same author. 



