OF NATURAL HISTORY. 255 



when left to Nature, never brought forth white flowers, nor the 

 white kind red flowers. 



Upon this experiment we have to remark, i. That nothing is 

 more dangerous, or more fallacious in philofophy, than the afliimp- 

 tion of general pofitions without an accurate inveftigation. The 

 Dodtor advanced, for example, that the red and the white lychnis, 

 when in a natural fl:ate, never change their colours. This pofition 

 is neither capable of admiflTion nor denial ; becaufe no experiment, 

 nor inquiry, feems ever to have been made on the fubjedt : Yet it is 

 affumed as a premife to the conclufion, that the change of the white 

 into a red lychnis was occafioned by the influence of the red male 

 upon the white female. 



2. That hybrids, or mules, uniformly participate of both the fpe- 

 cies or varieties by which they are engendered. A jack-afs and 

 mare never produce a fimple afs or horfe, but a mule, or mixture of 

 the two. It fhould feem, however, that this red lychnis transfufed 

 its own individual qualities, without allowing a fingle particle of the 

 female to appear. This is contrary to every analogy. If the change 

 had originated from fexual commixture, the progeny ought not to 

 have been completely red, but pied, or a mixture of red and white. 

 To whatever caufe, therefore, this change may be attributed, it can 

 never be afcribed to any thing analogous to generation. 



3. That colour is a delicate and flu6tuating quality. It depends 

 fo much on light, air, health, and perhaps fome unknown caufes, 

 that botanifts, with great propriety, have rejeded it as a fpecific 

 character. Sufpedling that caufes of this nature might change the- 

 colour of the white lychnis under confideration, I examined the 

 condition of fome plants then fubjeited to the fame trials in our 

 Botanic Garden. The flowers both of the red and white lychnis- 



