4-12 MR. B. CLARKE ON THE STRUCTURE AND AFFINITIES 



flower no positive difference is apparent except in the carpels being right and left of the 

 axis, instead of antci-ior and posterior : and in its deficiency of floral envelopes, it may, as 

 far as relates to the corolla, be compared to the Stilbacece, near allies of VerbenacecB, the 

 female flowers of which are apetalous. 



The position of the Baiidece, therefore, ajipears undoubtedly to be with the mono- 

 petalous families ; and however contrary to the rules of affinity it would have appeared 

 only a few years since, this makes it in some degree probable that CulUtrichacece should 

 also be stationed there, and near JBatidece, as the ovary has precisely the same structure, 

 being 4.-celled, and consisting of two carpels, right and left of the axis. In Callitriche, 

 however, the ovules are suspended, amphitropal, with the raphe next the placenta, in which 

 it coiTcsponds with Boraginea, or very nearly so * ; the seed is albuminous, and the 

 cotyledons are very short, as in Stilbacea, with which it further agrees in its anther, which 

 closely resembles that of Campylostachys in shape, and in the peculiarity of the two cells 

 being confluent at the apex, so that at the time of dehiscence it appears as if it had but 

 one cell. 



Callitriche may be compared with Elatinece, especially if the structure of the ovary of 

 Tetradiclis can be considered as analogous to that of Callih^iche ; but Bcdis does not 

 appear to have any affinity with the polypetalous families, except that the long claws 

 of its petals are like those which occur in Caryophyllacece. Callitriche differs from Eu- 

 phorbiacece in its amphitropal ovules; besides which, hermaphrodite flowers are not 

 uncommon. 



2. Note on the Structure and Affinities of Vochysiacese. 



One of the most remarkable characters in this family is that the carpel when single is 

 posterior, as in Erisma ; the single cell of the ovary is on the opposite side of the flower 

 to the stamen, which is anterior, and in direct relation with the spur of the calyx, which 

 belongs to the posterior sepal (fig. 28). This character I believe completely separates the 

 Vochysiacece from the epigynous families, and especially from the Onagracece, with which 

 tliey have been compared, as in that family the carpel when single is anterior ; nor is it 

 known to be always posterior in any of the epigynous families. 



If the position of the single carpel is in the present instance taken as a guide to affinity, 

 this family should be associated with one in which the flower is in some degree irregular, 

 with the single carpel posterior ; and these characters occur in JDaphnacece, of which the 

 Vochysiacece may be a polypetalous form t, standing in the same relation to the Chi^yso- 

 halanece and Leguminosce, and especially to the former, as Daphnacece to Proteacete. 



The relation, however, between Vochysiacece and Dap)hnacece is rather one of analogy 

 than affinity, although perhaps there is no character to separate them except the absence 



* In Borat/inece the ovule is scarcely suspendetl, being attached near or at the base of the cell, at the inner angle ; 

 but the raphe of the amphitropal ovule is on its under surface- — which is, doubtless, equivalent to the ovule being 

 suspended with the raphe next the placenta. 



t In Pimelea and other genera of Dajt/inacea: there is a decided tendency to irregularity of the flower of the same 

 kind as in Grevillea among ProteacetB, with the exception that the carpel is always posterior instead of anterior, and 

 it is not uncommon for flowers of Pim^/ea to have but one stamen anterior, as \n Erisma (vide Ann. Nat. Hist. 

 2nd ser. vol. xi.). 



