CLASS DODECAN'DRIA IMPERFECT. 125 



number of all the parts of the flower, especially of 

 the pistils, varies according to the richness of the 

 soil in which the plant has grown ; and it was for 

 this reason that I wished to bring home specimens 

 from different situations. The branch we have 

 just examined, was from the thatched roof of the 

 cottage, where I suppose the soil was richer than 

 on the top of the wall. 



Houseleek was formerly planted on the tops of 

 houses, from a superstitious idea that it was a pre- 

 servative from lightning and thunder. 



EDWARD. 



I am afraid that I should not have been able 

 to find out the name of this plant, if you had not 

 told me. 



MOTHER. 



It is unsatisfactory to examine the plants which 

 are placed in this class, because the character itself 

 is imperfect. You must always recollect, that 

 eveiy botanical arrangement, or System, as it is 

 called, must be defective ; because the acquaintance 

 with plants, even of the best botanists, is very far 

 from being complete. The method of Linnaeus 

 is, upon the whole, one of the best; but it has 

 some faults, and the imperfection of this eleventh 

 class is one of them; for the class has no found- 

 ation in nature, and has been formed merely for 

 convenience ; whereas the grasses, and some othei 



