On the Affinities of Ceratophyllacea. 47 



native of Cayenne, the other of S. Carolina, Louisiana, &c. 

 Both species have the habit of Brasenia as to the flowers and 

 floating leaves, and that of Ceratophyllum in the filifonnly 

 dissected submersed foliage. Brasenia and Cabomba are very 

 nearly allied ; the principal difference consisting in the more 

 numerous stamens and ovaries of the former, while in the 

 latter the ovaries are reduced to two, three, or four, and the 

 stamens are (as in Ceratophyllum) only twice the number of 

 the sepals. The ovaries of both bear from two to three ovules, 

 but commonly only one or two seeds are perfected. The 

 indehiscent carpels are crowned with the persistent styles, as 

 in Ceratophyllum. 



In order to bring to view the real affinities of Cabombaceae, 

 it is necessary to avail ourselves of the important characters 

 furnished by the seed, the true structure of which (particularly as 

 compared with allied orders, in the direction of the radicle, &c.) 

 has been almost wholly overlooked.* The seeds of Cabomba 

 and Brasenia resemble each other very closely. They are sus- 

 pended in the pericarp, one above the other, when two are present; 

 and a rather large roundish spot (similar to that on the seed 

 of Nelumbium) is observed at the extremity opposite the hi- 

 lum ; which, as no raphe is perceptible, may be inferred to be 



2. Cabomba Caroliniana ; foliis natantibus elllpticis lineari-oblon- 

 gisve ; floribus albis ; ovariis 3 — 4. — C. Aubletii, Michx. ft. 1, p. 206. 

 Nectris peltata, Pursh, ft. l,p. 239 (excl. syn.). N. aquatica, Nult. gen. 

 l,p. 230; Ell.! hot. 1, p. 416, non Willd. — Hab. In aquosis S. Caro- 

 lina? ! Georgise ! et Louisiana; ! 



I have seen flowers of this species with only two sepals, and two 

 petals. Pursh, who considers the North American plant the same as 

 that of Aublet, has, nevertheless, taken the unwarrantable liberty of 

 changing the specific name. 



* Fine dissections of the seed, both of Cabomba and Brasenia, 

 drawn by Richard, are published in the plates of the Did. dcs Sciences 

 Naturelles, from an examination of which their true structure may be 

 deduced. It is worthy of notice that the portion of the embryo which 

 Richard calls the gemmule in Ann. du Muscvm 17, p. 230, (. 5, Jig. 22 

 and 23, he here more properly considers to be the radicle. 



