INTRODUCTION. Ixiii 



The number of the Ovula (that is to say, whether they are definite or 

 indefinite) is frequently an important difference, as, for example, between 

 Campanulacese and Compositae, Goodenoviae and Scaevoleae ; but while 

 I think considerable value usually attaches to this, it must not be forgotten 

 that there are exceptions to it in several instances, especially in Caprifolia- 

 ceae, if Hydrangea really belongs to that order, and Fumariaceae and Cru- 

 ciferae. — The position of the ovula is much more essential than their num- 

 ber, and may be considered as one of the most valuable forms of structure 

 that can be taken into account. It is uniform in Compositae, Valeria- 

 neae, Umbelliferae, and others, and it constitutes an absolute distinction be- 

 tween Artocarpeae and Urticeae; but in Sanguisorbeae, Pedalinese, and Sty- 

 raceae, both erect and suspended ovules co-exist ; this union of the two 

 positions occurs in a most remarkable degree in Penaeaceae ; and among 

 Violaceae, the genus Conohoria offers, according to M. A. St. Hilare, (PL 

 Usuelles, No. 10,) an instance of three kinds of direction in as many spe- 

 cies; in C. Lobolobo, the ovula are ascending; in C. Castaneaefolia, they 

 are suspended, and in C. Rinorea one is suspended, one ascending, and 

 the intermediate peritropal, or at right angles with the placentae. — The 

 situation of the foramen of the ovulum is a circumstance which should 

 always be taken into account, because it indicates with certainty the future 

 position of the radicle, which it is of first rate importance to ascertain, but 

 which will be more properly spoken of in considering the value of distinc- 

 tions drawn from that source. 



The ripened ovarium is the Fruit. The differences in its structure 

 are of the same nature as those of the ovarium, and need not be repeated. 

 Its texture and mode of dehiscence are the principal sources of distinctions, 

 but they perhaps deserve as little attention as any of which botanists make 

 use. It is true that the fruit of all Grossulaceae is baccate, of all Labiatae 

 indehiscent, and of all Primulaceae capsular; but Marcgmaviaceae, Mela- 

 stomaceae, Myrtaceae, Ranunculaceae, and Rosacea?, and a crowd of other 

 orders, contain both baccate and capsular, dehiscent and indehiscent 

 genera. 



The characters obtained from the position of the Seed are of the same 

 value as those from the position of the ovula ; in addition to which, the 

 peculiarities of the testa are made use of. In some Monocotyledonous or- 

 ders, as Asphodeleae and Smilaceae, the texture is employed as a mark of 

 distinction ; its being winged or otherwise distinguishes Meliaceae from 

 Cedreleae, and the presence of a fungous swelling about the hilum is a 

 good characteristic of Polygaleae. Linnsean botanists make a distinction 

 between naked and covered seeds, attributing the former character to La- 

 biatae, Boragineee, &c. ; but the sense in which they use the term is so 

 manifestly erroneous, that botanists were at one time led to believe that 

 no such things as naked seeds existed. It is now, however, known, from 

 the accurate observations of Mr. Brown, that certain tribes of plants do 

 exist in which the seeds are really naked, that is to say, susceptible of im- 

 pregnation and maturation without the intervention of any pericarpial 

 covering. These are Coniferae and Cycadeae, orders exceedingly remarka- 

 ble in other respects, especially in the structure of their vascular tissue. In 

 consequence of these peculiarities, they have been distinguished by A. 

 Brogniart as a class of the same dignity as Dicotyledons and Monocotyle- 

 dons. Without assenting to this proposition, to which I think there are 

 great objections, it is impossible to doubt that the naked seeds of these or- 



