including a new Ajrangement of Phanerogamous Plants. 83 



lumbium. 2. Hydropeltidese (ovules two, one above the other). 

 3. Ranunculaceae. 4. Monhniacese. 5. Lauracese. 6. Ana- 

 cardiacese. 7. Coriaria (PI. II. figs. 1 & 2). 8. Malpighiacese 

 (in those genera in which the funiculus is next the dorsum of 

 the cell). 9. Celastraceae. 10. Ternstromiacese (ovules two, col- 

 lateral). 11. Ebenaceje (ovules two, collateral). 12. Icacinese. 

 13. Loranthacese (PL II. fig. 3, and see also Part III.). 14. Plum- 

 baginese. 15. Illecebracese. 16. Chenopodiacese. 17. Ama- 

 ranthacese. 18. GemoZowzc! (ovules two, collateral) . 19. Cyril- 

 lacese. 20. Helwinyia. 21. Aucuba. 22. Cinchonacese (PI. II. 

 fig. 4). 23. Tetragoniaceae. 24. Calyceracese. 25. Dipsacus. 

 26. Onagrarise (ovules three or four). 



The campylotropal ovule in which the radicle of the embryo 

 subsequently formed is turned towards the placenta, as in 

 Amaranthacese, is a character which deserves especial atten- 

 tion, if it is equivalent to that of the raphe averse in the pen- 

 dulous anatropal ovule, and that it is so, Staiice and Plumbago 

 seem to prove. That Plumbago is a genuine instance of the 

 raphe averse there appears no reason to doubt, because it is 

 constantly so, — the i*aphe being always on the side of the ovule 

 which is directly away from the funiculus, the latter coiling 

 round the edge of the foramen to join it. And if its occur- 

 rence in the ovule of Plumbago is admitted, then there can 

 scarcely remain a doubt of the correctness of the inference in 

 question, because in Gom2)hrena and Philoxerus, where the ovule 

 is equally suspended as in Plumbago, the foramen and subse- 

 quently the radicle are always next the funiculus. It may be 

 added also, that in Scleranlhus annuus the ovule as regularly 

 curves away from the funiculus, as in Gomplirena it curves 

 towards it (see also Part III, and the accompanying figures) ; 

 and this question is almost set at rest, when it is considered that 

 no distinction which is absolute exists between anatropal and 

 campylotropal ovules, as in Trianthema the ovules are simply 

 campylotropal, but in the nearly allied genera Galenia and Tetra- 

 gonia a short I'aphe is present, although the ovule is curved as 

 in the former case*. 



2. Ovule pendulous with the raphe lateral. This has beea 



* Since tlie above was written, I have ascertained that in Atriplex the 

 same inversion of the ovule takes place as in Euonymus and Raniinculacese 

 when it is single and pendulous; thus in A. angustifolia the seed is erect 

 and the cotyledons next the jijaceuta, being on that side of the ovary to 

 which the short funiculus is adherent; in A. laciniata the seed is attached 

 above the middle of the wall, so that being vertical, the cotyledons are on 

 the upper surface, and the radicle undei'neath curving up so that its extre- 

 mity reaches to the hilum; and in Halimus pedunculatus (olim A. pedun- 

 eulata) the inversion is complete, the seed being pendulous and the cotj'- 

 ledons turned away from the funiculus. 



6* 



