ON THE NATURAL SYSTEM. 



5 



equal in value to three. I also feel confident that the position of the single carpel in families in which it is unknown may, in nearly all instances, be with moderate certainty calculated from 

 their structure and affinities, so that outlines of Subdivisions or Sections may be thus formed which in most instances will I believe be found natural. It might be supposed for instance that the 

 Second Section of Table IV. could not be regarded as Procarpous, but as there are included in it five families, besides Ranunculacece, in which the flowers are sometimes more or less irregular, 

 viz., Pyrolacea, Ericaeea, Ochnacetz, Pittosporacece, and Marcgraviece and in which consequently the single carpel would be in all probability anterior, it becomes at least theoretically probable that 

 it is so, Ceratophyllacece and Nelumbiacece being the only exceptions as yet known. 



In the Vertebrata taking the alimentary canal as the axis, the female sexual organs or their ducts are either posterior or anterior and the latter section are the more highly developed, and 

 although this difference in the position of the male organs is less marked, yet it is already admitted as a guide to affinity as abdominal testes are acknowledged to indicate an inferior type. In 

 Mammals the uterus and vagina are situated quite anterior, while the corresponding parts in Birds, Reptiles, Amphibians, and Fishes are posterior or show only a tendency to become lateral, so that 

 Mammals might by the position of the uterus be separated from the other sub-classes of Vertebrata.* This rule is in a limited degree applicable to plants, the sections formed, being, as far as at 

 Present known, smaller and less distinctly marked. Differences of the same nature occur also among the In vertebrata, but how far they would form natural sections may be a question. 



Value of Other Floral Characters. 



Having endeavoured to show that the epigynous is the lowest form of ovary among the Exogens, the next step naturally is to enquire which is the highest form, so as if possible to separate 

 one or more Divisions from the Perigynous or Hypogynous families which might be regarded as the highest developed form of plants. On the principle that the Epigynous are a lower Division 

 of plants than the Hypogynous, it might be expected that the Perigynous Division would prove to occupy an intermediate station between the two, but the families regarded as Perigynous are so 

 intimately blended with the Hypogynous (the character itself being scarcely definite), that any separation of them appears inconsistent with a due regard to the most strongly marked affinities. 



The ovary which differs most from the epigynous it will probably not be doubted, is the apocarpous, and this is one of the characters which is regarded as showing a higher degree of 

 organization, as none of the apocarpous families are I believe near the Epigynous Division. To this rule however there are certainly such exceptions as to make it only of secondary value, but 

 the Chloranthal Division, Table III., which I believe to be the lowest form of Perigynous and Hypogynous plants, contains no strictly apocarpous family, and in fact scarcely any apocarpous 

 genera, while the Platano-Proteal Division Table VI. contains more than any other. 



The ovule pendulous with the raphe next the placenta, especially if occurring in conjunction with the single carpel anterior, is regarded as a character always showing a- comparatively 

 higher degree of development, and such families are always placed foremost in the Subdivisions of the Tables. And where the position of the single carpel is unknown, the ovule 

 pendulous with the raphe next placenta is still taken as a guide, so far as to place such families in the most advanced positions in the Subdivisions as far as is consistent with other affinities which 

 very rarely interfere with such arrangements. From the very rare occurrence of this character among the Endogens as before noticed, it might be expected that it would prove applicable to a 

 considerable extent in showing the position of many of the Exogenous families. 



The occurrence of irregular flowers where they are irregular in the highest degree, especially if the ovary is reduced to one carpel and that anterior, and the raphe where the ovule is pendu- 

 lous is next the placenta, is beyond doubt also a character indicating a comparatively higher degree of development even though it may not extend through the whole of the family, and it may be 

 a question if there is any material exception to this peculiarity of structure as a guide to affinity as far as regards Subdivisions. 



On these principles the Proteacece and Leguminosce and their allies, in which the apocarpous ovary occurs in conjunction with the irregular flower,f and the didynamous families especially the 



* In the Chamaeleon, which is among the higher developed Beptiles, the two oviducts combine and form a single tube as in Fishes, but the two ducts do not form one canal at any part, and they open by separate orifices. At their lower 

 termination they form by uniting together a double tube situated behind the large intestine. In the Tortoise and Frog the oviducts are very large and lie on the sides of the large intestine but open behind it, being separated at their termination 

 only by a very thin septum, so that their orifices are close together, which becomes more obvious if they are distended. In both of them the urethra is in front of the intestine, so that the orifices of the oviducts, which in the form of vagina are 

 in close relation with it in the Mammals, are widely separated from it. 



f Leguminosce are, in the position of the raphe in suspended ovules, more highly developed than Rananculace<2, ( V. Table IV.) 



