350 Tricotylous Races. 



of commercial seed the application of this term would 

 not, of course, be justified; but in my cultures, which, 

 almost without exception, were started by the selection 

 of tricotylous individuals, this term is obviously fully 

 justified. Moreover, in this way, the word dicotyl is 

 left to its pure systematic signification. And, as in some 

 other cases, atavism is here seen to be an oscillation be- 

 tween two empirically known extremes.^ 



So far as my experience goes, tricotylous seedlings 

 are much commoner amongst cultivated species than 

 amongst wild ones, and even amongst the latter they 

 usually occur from seeds saved in botanical gardens and 

 ^'ery seldom from those collected in the field. Amongst 

 cultivated plants, again, they are commoner amongst spe- 

 cies which are grown on a large scale than on a small 

 one. 



Thus I have obtained my intermediate races from 

 amongst the former, partly from agricultural and partly 

 from horticultural species. As an instance of the former 

 I may mention the hemp and of the latter the snap-dragon 

 and the wild poppy. But besides these I discovered a 

 rich tricotylous race in my Oenothera hirtella, an entirely 

 new species which was found quite by chance in my 

 garden. It is not impossible that cultivation on a large 

 scale favors the origin of new races. 



A profusion of instances of some anomaly in a species 

 either in the field or in cultivation, suggests the presence 

 of an intermediate racer a scarcity, however, that of a 

 half race, as is especially well known to be true in the 

 case of fasciations. Tricotyly conforms strictly to this 

 rule. From seeds obtained from the trade or by exchange 



^ See page io8. 

 * See page 32. 



