ORIGIN OF ENDOGENS FROM EXOGENS. 503 



the line of union being just indicated by the two delicate epi- 

 thelioid layers. 



I would also invite the reader's attention to the series of 

 figures by Hegelmaier tracing the development of the embryo of 

 Trop&olum. However, neither he nor Irmisch appear to have 

 noticed the four divisions making the quadrants, which are 

 so far distinct as to have a minute hole just over the tip of 

 the radicle. Consequently Hegelmaier's figure # is evidently 

 wrong. Sir John Lubbock rightly describes it f. Irmisch, 

 however, shows the separation as seen in one direction J. 

 Trecul § has described the basipetal origin of the lobes in Tro- 

 p<Bolum, and explains how the peltate blade is formed by a basal 

 outgrowth connecting the lowermost lobes, as occurs in Nelum- 

 bium and Umbilicus. 



On examining the embryo of Tropaolum majus for the pur- 

 pose of studying the nature of the pseudo-coleorhiza and compar- 

 ing my observations with Irmisch's figures, there seemed to be 

 several points which suggested an ancestral aquatic habit. Thus 

 it will be seen || that the first series of leaves formed on germi- 

 nation resemble those of Nymplicea and Sagittaria. The first is 

 a subulate phyllode ; the second has a linear spathulate blade ; 

 the third is subhastate, or somewhat suggestive of the adult 

 leaf of Menyanthes trifoliata ; the next being lobed and peltate IF, 

 T. tricolor. The orbicular leaf of T. majus suggests a similar 

 origin to that of Cabomba, Victoria, Nelumbium, and Hydrocotyle; 

 for in addition to the above features the stem of Iropceolum 

 has regularly formed air-canals, as in true aquatic plants, and 

 water-pores over the nerve-endings as in endogens. Lastly, 

 Tropceolum has the marsh-plants Limna?itJies and FloerTcea as its 

 nearest allies. It will be seen hereafter, when comparing the 

 germination in the Nyrnphceacece with that of Sagittaria and the 

 Aroidece, how that the sagittate form is pre-eminently charac- 

 teristic of both an aquatic and an endogenous type of leaf. 



Vergl. Unters. Entw. dicot. Keime, Taf. ix. fig. 28. 



t On Seedlings, i. p. 297. 



\ Beitr. Vergl. Morpb. Pfl., Taf. v. figs. 17, 19. 



§ "Sur la Formation des Fouilles," Ann. des Sci. Nat. 3 fie'r. xx. 1853, 

 P. 235, pi. 23. figs. 85-91 (Tropaolum), pi. 25. figs. 153, 154 (Xelumbium). 



[| Op. cit Taf. t. figs. 6-9. 



1 Richardia furnishes the intermediate condition of having sagittate- peltate 

 leaves. These are aerial, but represent the aquatic form. Further observations 

 °n the Aroidea will be made hereafter. 



