[ 8? ] 



dcner, who propagates without any variation of kinds, for 

 centuries together, the different varieties of beans and 

 peafe ought certainly, if adverted to, to have excited a 

 doubt at lead of the univcrfality of the rule. 



In the fame manner that I obtained, as above, a new va- 

 riety of turnips, it might poflibly happen that fome new 

 varieties of potatoes might in fome cafes be obtained from 

 feeds. For if the varieties of this clafs of plants naturally 

 admit of mixture with each other, (which however I doubt) 

 and if feveral kinds of potatoes mould happen to grow in a 

 field mixed together, two or more of thefe forts blended to- 

 gether might produce a mongrel breed, participating of 

 the qualities of the parent ftocks ; but as to new varieties, 

 obtained from feeds gathered from folitary plants, I have 

 met with no fact that gives reafon to expect them. 



The only cafe that occurs to me juft now as feeming to 

 confirm the doctrine of feminal varieties, is that of fruit 

 tree, which I only mention that it may not feem to be 

 overlooked. But I have met with no facls that tend to 

 prove that the generally received notion on this head is not 

 hypothetical, and have found fome that feem to prove that 

 it is. It is in general very confidently afTerted, and as 

 generally believed becaufe of that afTertion, that the feeds 

 of a grafted apple do not produce trees bearing fruit of the 

 fame kind with themfelves. This is fo generally believed 

 in Europe, that I have never heard of an experiment being 

 made with a view to prove it : but that very good apples 

 are produced from feeds without grafting, is certain by the 

 practice in America, where that operation is entirely laid 

 afidc, and where the fame kinds of apples are frequently 

 found on different trees in the fame way as among the 

 grafted trees in Europe i though doubtlefs, where the feed* 

 arc allowed to fow themfelves at random, there muft be 

 G 3 much 



