N A TU R A 1 . CLASS! ¥ 1 C A T I O X . 



119 



state of the unripe germen, and of the ripe fruit, witliout re- 

 marking, that the former contains several ovula, whilst the 

 latter always incloses but one seed. In the Siliculosae, as in 

 Crambe, Cakile^ and Myagrum^ we always find empty abor- 

 tive loculi. Is it further necessary to multiply examples, 

 by calling to recollection the common Snow-ball, and Hydran- 

 gea hortensis, where the abortive state of the sexual parts 

 affords opportunity to the evolution of the beautifully colour- 

 ed integuments of the flower ? 



178. 



A question naturally arises, in what way we may avoid the 

 mistakes to which such abortion may lead. For this purpose, 

 we propose the following means. 1. We must examine the or- 

 gans in all their proportions, even in those that are most differ- 

 ent. It cannot but happen, that during such examination a re- 

 turn to the symmetrical or natural structure will be observed. 

 These symmetrical proportions are often the natural ones, • 

 and art has merely produced the abortive state, by evolving 

 one organ at the expence of another. The Snowball of our 

 gardens bears perfect flowers when it grows wild. It is fre- 

 . quently art itself, or a luxuriant growth, which produces 

 this return to the original structure. As, for example, the 

 appendages to the filaments of Sage are abortive filaments, 

 the plant, like other Labiatse, should properly have four fila- 

 ments. In fact, we observe, that, in many species of this ge- 

 nus, the oblique processes of the filaments carry anther-shaped 

 bodies; and in Salvia gluiinosa^ we find these, during moist 

 summers, passing into true anthers. When we compare with 

 these Stachytarpheta and Westringia^ we see very distinctly 

 the two abortive filaments, which also, in the last named ge- 

 'nus, bear empty anthers. 



But in every case it is important to observe the organs at 

 their first appearance, whilst they are as yet unrestrained in 

 their structure. Thus, we must examine (and this is one of 

 the most important rules for the botanist) the loculi of the ova- 

 rium in its unripe state, in order to determine respecting the 

 true nature of the ripe fruit. When we examine a spine in 



