XVII SUMMARY AND CONCLUSIONS 605 



the gametophyte and embryo show characters that are more 

 Hke those of the Ferns, and the exact position in the system 

 of Isoetes must still remain somewhat doubtful. 



The Angiosperms are in all probability all members of a 

 common developmental series, but just what is their relation to 

 one another and to the other vascular plants is not so evident. 

 It is usually held that they have been derived from the Gymno- 

 sperms through the Gnetacese, but it has also been suggested 

 that one or both of the divisions may have originated directly 

 from the Pteridophytes. Attention has been called more than 

 once to the close resemblance between the embryos of the Fili- 

 cineae and those of typical Monocotyledons, and this is especially 

 the case in Isoctcs, where, in addition, the. structure of the 

 mature sporophyte is much like that of the Monocotyledons. 

 It is possible that the surrounding of the sporangium by the 

 base of the sporophyll may be the first indication of the ovary 

 of the Angiosperms, but as this applies to the microsporangia 

 as well, much stress cannot be laid upon it. It is quite as easy 

 to trace back the embryo-sac of the Angiosperms to the macro- 

 spore of Isoetes as to the embryo-sac of the Gymnosperms ; and 

 when the great similarity between the sporophyte of the former 

 and the Monocotyledons is considered, the probability of the 

 origin of the latter from aquatic or semi-aquatic ancestors 

 resembling Isoetes is certainly considerable. 



The essential similarity in the structure of the embryo-sac 

 in all Angiosperms yet examined, as well as the structure of the 

 flower, makes it almost inconceivable that the two branches, 

 Monocotvledons and Dicotvledons, could have arisen from dif- 

 ferent stocks. Strasburger's suggestion that the Dicotyledons 

 were derived directly from the Gymnosperms, and that the 

 Monocotyledons are a reduced branch of the former, is open 

 to objections both on morphological and palseontological 

 grounds, and we believe that the evidence we have at present 

 points to the Monocotyledons as the more primitive of the two 

 divisions of the Angiosperms, from which later the Dicotyle- 

 dons branched off. It is true that the researches of the past ten 

 years (Coulter (4)) show that there is less tmiformity in the 

 structure of the embryo-sac than was supposed to be the case; 

 but there is no question as to the essential similarity in struc- 

 ture in all Angiosperms. It is also becoming evident that the 

 dicotyledonous habit may have developed more than once. 



