including a new ArirangeMenl^of Phtinm'dge^o^i Plants. 44&^ 



' 'PtANTAGiNKJE. Iti Liltorella the axils are three-fl<)wei'ed,4he 

 central being axis to the two lateral ; in this family the raphe is 

 next the placenta, but it is here feebly marked, and I can only 

 say that it is variable, being rarely, but I believe occasionally, 

 anterior, from which it would appear that the fertile carpel is for 

 the most anterior or lateral, thus approaching Dipsacese in this 

 character. 



IllecebracejE. In Illecebrum and Herniaria the stigmas are 

 all anterior and posterior, the ovary consisting of two carpels 

 united by their margins ; and the seed is erect, having the radicle 

 always curved down on the posterior side, thus agreeing with 

 Atriplex and Chenopodium Bonus Henricus both in the position 

 of the stigmas and of the cotyledons and radicle. In Paronychia 

 the examination is difficult in consequence of the diminutive size 

 of the ovary ; in two or three instances however the funiculus 

 was evidently anterior (in P. capitata), and in others apparently 

 so, no variation being observed. 



CHENOPODiACEiE and Amaranthace^. In Chenopodiacese 

 and Amaranthacese the ovary has the appearance of consisting 

 of carpels united by their margins as in Polygonacese, Alsineae, 

 and IllecebraceEB, the ribs of the ovaries, if any are present, 

 being opposite the stigmas, and when the stigmas are two being 

 equally marked on both sides *. In Beta, &c. the funiculus ad- 

 heres to the side of the ovary, and this adhesion is for the most 

 part anterior, rarely right or left ; but if the funiculus is free, as 

 in Rhagodia, it still has the same position, being also in either 

 case opposite a stigma when the ovary is digynous. As the 

 funiculus in these instances is the remains of a central placenta, 

 the position of the ovule shows a tendency to the production of 

 the single carpel for the most part posterior as in Piperacese, and 

 surely this is placed beyond doubt when it is recollected that in 

 Atriplex and Chenopodium when the seed is erect the cotyledons 

 are anterior. 



In Gomphrena the funiculus is lateral and the stigmas right 

 and left, which may be regarded as an approach of Amaran- 

 thaceae to Nyctaginese, and also to Polygonaceaj, as in Polygonum, 

 when dicarpous, the carpels are all right and left. 



Bpacride^. The ovary of Acrotriche cordata is not unfre- 

 quently one-celled, and the position of the carpel is then va- 

 riable, being frequently posterior. 



CoMBRETACEiE. The style of Combretum is more or less ob- 

 lique in its direction, as also the ovary, and the obliquity fre- 



* To this it may be added, that the ovary of Basella evidently consists of 

 three carpels united by their margins, three smaller ribs alternating with three 

 larger, and the latter being continuous with the stigmas ; the position of the 

 ovule is uniform, and I believe corresponds with that of lllecehrum, &c. 



