98 Flowers and their Pedigrees. 



gardens it is usual to plant one row of 'barren' 

 flowers for every three rows of ' fertile ' ones, leaving 

 the insects to do the rest. At present none of these 

 varieties can be said to have developed into what old- 

 fashioned botanists used to call ' a good species,' for 

 fertile cross-breeds can still be readily produced 

 between them all by artificially fertilising the pistils 

 of one sort with pollen taken on a camel's-hair brush 

 from the stamens of another. The possibility of 

 fertile hybridisation in such a manner shows that the 

 plants have not long diverged from the common 

 central stock. But after they have long been exposed 

 to varying circumstances and acted upon by natural 

 selection, they will probably become so different from 

 one another in a variety of small particulars that the 

 hybrids will no longer prove fertile, owing to the want 

 of sufficient similarity between the respective ancestral 

 lines. Perfect fertility is only possible between indi- 

 viduals which %till retain all the principal traits of a 

 common ancestral form. Curiously enough, one ex- 

 isting variety, the Himalayan strawberry, has actually 

 reverted to the primitive yellow flowers of its cinquefoil 

 allies. 



On the other hand, if the strawberries ever really 

 live down the white potentillas, so that the latter race 

 dies out altogether, then the distance between the 



