including anew Arrangement of Phanerogamous Plants. 451 



the fertile carpel in Opercularia* fully corresponds "with that df 

 Aucuha in being more frequently posterior. 



Caprifoliace^. In Abelia the position of the one-seeded 

 and only fertile cell is variable, and more frequently posterior 

 than the fertile carpel of Viburnum, and as in that genus the 

 raphe is lateral. (PI. XIII. fig. 4.) 



CornacejE and ALANGiACEiE. The ovary oi Aucuba japonica 

 is probably not compound, as there are no remains of an abortive 

 carpel ; when fully developed the stigma is unilateral and almost 

 horizontal, being stigmatic and somewhat grooved on its upper 

 surface, and smooth and rounded beneath ; the ovule is not 

 attached to the apex of the cavity, but to the upper part of 

 that side (immediately below the apex) which is away from the 

 stigma, and the placentation consequently varies with the posi- 

 tion of the stigma, being anterior when it is posterior and vice 

 versa (PI. XIV. fig. 15). In Marlea the two cells are for the 

 most part unequal, the smaller one being often much reduced in 

 size, and the position of the larger cell corresponds nearly with 

 that of the single carpel of Aucuba in being more frequently pos- 

 terior or lateral. 



GarryacEjE. The ovaiy of Garrya has all the appearance of 

 consisting of two carpels united by their margins, as the ovules 

 are attached to opposite sides of the ovary near the apex, their 

 attachment alternating with the stigmas. It may therefore be 

 regarded as having the same structure as that of Helwingia with 

 retracted dissepiments, and with that genus as being a near ally 

 of Santalacese. It may further be compared with Lauracese and 

 also Daphnacese, if in the latter family, when dicarpous, the pla- 

 centation should prove to be parietal. 



Proterocarpous Exogens. 



The connexion between the Polypetalous and Monopetalous 

 Orders is here not always so distinct as in the Heterocarpous 

 Division, and whether Solanacese are a monopetalous form of 

 Papaveracese is rather a question, but if arranged with Polemo- 

 niacese they would in Table III. be placed as before with their 

 nearest allies. 



Fumariace^. In Fumaria officinalis one carpel is larger than 

 the other, and to it the seed is always attached and almost ex- 

 clusively so, the hilum being oval (the junction between the two 

 carpels is however not obvious externally, but is distinctly seen 



* The axis is here compound, each capitulum consisting of several cen- 

 tres of growth indicated by small whorls of partially adherent ovaries. The 

 raphe is for the most part anterior, being lateral or posterior in only two 

 or three of six instances. 



