16 INTRODUCTION. 



characters. I refer to these authors for exam- 

 ples so often met in gardens ; but I have chief- 

 ly attended to this in the woods and fields 

 where quite spontaneous. 



The process is by the seedlings being some- 

 what different from the parents, and thus evinc- 

 ing a deviation of typical mould, that may be, 

 or may not be, propagated again. If it is, this 

 soon assumes a permanence, becoming a per- 

 manent variety if the deviation is slight, such 

 as mere color of flowers, size of stem, leaves, 

 &c.; but becoming a New Species! if at last 

 several deviations are permanently combined. 

 A tendency to such deviations is sometimes met 

 even in the various annual shoots of the peren- 

 nial plants, or shrubs and trees, that are not al- 

 ways alike to those of the preceding years. 



The specific deviations which I could men- 

 tion are numberless, this work will be full of 

 them, as all new species are in fact such per- 

 manent deviations of growth, unless they are 

 widely different from all former species. The 

 oldest type of the species may probably be 

 found in the most common with most numerous 

 individuals, while those called rare or with few 

 individuals as yet must be the newest in order 

 of time. 



Generic deviations are more rarely observed 

 or noticed, because less evident and very slow- 

 ly produced, or not so easily propagated ; but I 

 shall mention here some very striking instances 

 of them ; while many more are to be met with 

 in all the Genera, where the characters are 

 polymorphous, expressed by SO . . . OR . . . . SO, 

 3-5 fidis, 4-5 andris, 3-4 stylis, Capsule with 

 4 or 5 valves, one or many seeds, &c. 



1. I have seen in a garden a Tulip with 5 



