THE VARIABILITY IN CULTIVATION OF HARDY FLOWERING PLANTS. 225 



The hybrids it makes with the biennial V. Blattaria are nearly 

 as variable in colour, but are straggling and untidy plants ; but 

 the hybrids it makes with the perennial V. nigrum are less 

 variable in colour, being different shades of copper or orange, but 

 for dwarfness and abundance of flower are excellent for gardens, 

 though I have found them difficult to divide, and not lasting 

 many years. Other hybrids in the genus often occur, but are 

 less noteworthy. 



Interesting and remarkable mixtures, which, like those last 

 mentioned, seem barren, take place in my garden in the genus 

 Polemonium. The common Jacob's Ladder is not thought a 

 very ornamental plant, and as it is a profuse seed-bearer it 

 generally gives trouble in weeding, so it is not a favourite. But 

 amongst those seedlings which escape there are many hybrids. 

 Three species which have long been common in my garden are 

 (1) P. cceruleum, including the dwarfer form generally miscalled 

 P. Bichardso?ii, and the large Himalayan variety ; (2) P. reptans ; 

 (3) P. humile. I am not sure that No. 1 and No. 3 mix, but 

 hybrids between No. 1 and No. 2, and between No. 2 and No. 3, 

 are frequent. But the most remarkable hybrid in the genus is 

 a plant identical in flower with P. cceruleum, both in its blue 

 and its white form, which I have raised more than once from 

 seed collected in my garden of P. flavum. It is entirely barren, 

 and in this and two or three other ways an improvement on the 

 old type. Another change comes in P. cceruleum from luxuriance 

 of growth, and has often appeared spontaneously in my garden. 

 The leaf becomes bipinnate in form. This variety used to be 

 figured and sold as a distinct species named P. sibiricum. It 

 would be easy to enumerate other changes in plants probably 

 due to crossing. I will now mention a few cases in which 

 closely allied species produce from indiscriminate seeding a 

 doubtful series, some amongst which might pass by the name of 

 either parent. 



The common Harebell, Campanula rotundifolia, combines 

 with C. rhomboidalis on my rock beds, so that no line can be 

 drawn between them ; the seedlings appear fertile, and increase 

 the confusion. Linum flavum becomes very variable from seed, 

 not only in stature, ranging from 6 inches to 2 feet, but also in 

 breadth of leaf and in size of flower. I have more than once 

 obtained seeds and plants under the name of L. campanulatum. 



