LiyXEAN SOCIETY OF LONDOX. 59 



The problem of the true r.a(iir(> of the single colyledon of 

 Moiiocotvledou.s is still unsolved, but Coidter and Land's eoiitri- 

 butioii to tlie (juesliou (17) may go some way towards the solution. 

 They conclude from :i study of Ar/apanthas that cotyledons are 

 always lateral structures and that the single one is due to the 

 growth being concentrated into one rather than two priniordia. 



Kespecting the relative merits of an aquatic or geophilous 

 aucestry for Monocotyledons, the two views may be somewhnt 

 reconciled by regarding the earliest ones as neither markedly 

 aquatic or extremely geophilous — in fact, marsh plants witli stout 

 rliizomes. Some of thtir descendants have become completely 

 liydrophytic, others sharply geophytic, while others again have 

 retaken to the arborescent habit by fresh means. 



Ovale. The absencti of the orthotro])Ous o\uleiu Ihe lianales 

 nn'ght be advanced as an objection to our theory, especially as it 

 occurs in families considered primitive by the Englerians. This, of 

 course, is ou the assumption that orthotropy is primitive for 

 Angiosperuis. There iq nou reason to doubt this. It has been 

 shown for example in the case of Juglans (10. p. G28) and Uhnus 

 (9) that their ovules commence their development as anatropous 

 ones and gradually assume tlie orthotropous torm. Olher cases 

 of orlliotropy among Angiosperms deserve investigation from tliis 

 point of view. 



Embri/o-sac. On account of the meagre variation in the embryo- 

 sac of flowering plants, little, if an)', use can be made of it in 

 determining relationships within the group. That of Peperomia 

 was taken for a time as showing primitiveiiess, but this can hardly 

 any longer be maintaisied. The writer is inclineil t(; regard the 

 8-iiucleate sac as |)rimitive for Angios])erins and any departures 

 therefrom as derived (47. p. oSd). Thei'e is little hope of finding 

 among existing flowering plants a sac less reduced. 



.Since Welwitsclthi wwiX Gnetum i\,ve the only Gymnosperms which 

 do not form definite archegonia in their embryo-sacs, it is 

 tempting to compare their sacs with that of the Angiosperm. 

 That they form an interesting parallelism to that of the Howering 

 plant and are suggestive of the way the latter lias evolved from 

 the (xymnospermous sac nuiv be conceded; but that these 

 Gnetalean embryo-sacs are ])hyletically connected with that of the 

 Angiosperm is to me improbable. Pearson (39. p. 378) attempted 

 to connect them so, using that of Peperomia as a link. As 

 already mentioned this sac can no lon^i r be upheld as primitive. 

 Besides there are i;rave diiliculties in the wav of conneeling 

 ])hyleticall\^ the Gnetales with the I'iperales by using the highly 

 evolved genus, Peperomia, as an intermediary. 



Ewjlers Si/stem. Bugler's system, an elaboration of that of 

 Eichler's, which ousted largely but not wholly Bentham and 

 Hooker's founded on that of the Pronch school of taxonomy, has 

 had a considerable reign, and it is high tiuie for the sake of jjro- 

 gress in the study of the flower that it sliould no longer be blindly 

 followed, but critically examined with the view of the adoption of 



