C. Pei.lew and F. M. Durham 171 



Descriptiun of {t PI. XXVI, fig. 4. 



Plant less i-obust and smaller than Kewensis, very Horiferous. Many 

 superimposed whorls of flowers subtended by leafy bracts, erect flowering 

 stems, about 18 inches, peduncles thin. Leaves narrow ovate covered 

 with glandular hairs visible to the naked eye, varying in length from 

 1-celled stalked glands to 6-celled hairs with glands. Meal inside 

 calyx and on corolla tube as in Kewensis, slight meal at nodes. 

 Flowers large yellow, in size and colour as in Kewensis, but the petals 

 not so imbricate, and flatter. The anthers at th(! mouth of the tube 

 and the stigma at the same level; the anthers are usually shrivelled, 

 when pollen is produced it is small in (juantity, and examined micro- 

 scopically only about 40 per cent, of the grains appear good. The 

 average diameter of the grains is 5'6 jx, but there is great irregularity 

 in size and shape. 



The chromosomes of this plant were counted by Miss Digby, and 

 found to be 18 diploid and 9 haploid in number. In this it resembles 

 the original hybrid obtained at Kew, and this resemblance is borne out 

 by the marked sterility which it exhibits^. When used as a female, 

 either with its own pollen or with the pollen of other forms, this plant 

 is excessively sterile, but its pollen used on floribunda is quite effective. 



Self-fertilization has given the following results : 



1912. 7 flowers selfed, gave 2 plants ; 1 died before being 



described. 1 = f |. 



1913. Many flowers selfed, gave 1 plant = -\o^. 



1914. „ „ „ „ 2 plants ; 1 died before being de- 



•scribed. 1 = f]{. 



All the plants resulting from the self-fertilization of {| differ from 

 their parent and from one another ; they have characters in common 

 with the offspring of floribunda x i| ; they are smaller than the parent 

 but vary in size, more or less hairy, never mealy, with yellow flowers. 

 Three were thrums, and one pin. Of these plants : 



^1 has not yet been bred from. 



fl and 5i>_2 were perfectly fertile and gave a large number of plants 

 varying much, and with characters in common with the offspring of 

 floribunda X ||. Those characters which exhibit a wide range of varia- 

 tion, such as size, leaf shape, degree of hairiness, and sterility, evidently 

 depend on a large immber of factors which makes them extremely 



' This plant was shown to Mr Coutts and Mr Garrett of Kew, who said that it resembled 

 the sterile hybrid Kewensis first obtained at Kew. 



