xi] Double Primula 199 



if the plants are starved*. The female flowers are generally 

 normal, but sometimes extra petals are formed in them also. 

 Such female flowers have singular malformations which have 

 been often described, the most noticeable being an opening 

 of the ovary which causes the ovules to lie freely exposed. 



Of the hereditary descent of doubleness we know little, 

 but from what is known it appears that several distinct 

 systems are followed. 



In Primula Sinensis doubleness is an ordinary recessive, 

 singleness being completely dominant. The doubleness of 

 Primula is of a very unusual kind. It may be primarily a 

 petalody of the anthers, but I have never fully satisfied my- 

 self of this, nor do I know any critical observations on the 

 subject. When a large collection of Primulas is examined, 

 strains can usually be found, the members of which exhibit 

 the lower degrees of petalody in some or all of their flowers. 

 In the fully double flowers there is a complete series of 

 petals inside the normal ones and arising from them.. These 

 are formed as images of the outer petals, so placed that 

 their inner surfaces correspond with the outer surfaces of 

 the normal petals, and the two adjacent surfaces are struc- 

 turally both alike inner surfaces. The bizarre colour of 

 double Primulas is due to this circumstance. Inside the 

 corolla the stamens stand properly formed, and in their 

 normal relations. The appearance of these fully double 

 flowers is strongly suggestive of a delamination in the petals 

 themselves ; but as in those partially double flowers which 

 have one or two extra petals imperfectly formed and showing 

 their staminal origin, the petaloid tissue similarly faces out- 

 wards, it becomes impossible to distinguish any boundary 

 between the two phenomena. The morphological problems 

 which these facts create must be left to the expert botanist. 



The segregation of the single character from the fully 

 double is, in some families, clearly quite sharp, and doubles 

 always breed true when fertilised inter se. The inherit- 

 ance of the partial or petalodic doubles has been investigated 

 by Mr R. P. Gregory, and his experiments show that when 

 such a strain is crossed with a pure single, F^ is single, 

 and in F z singles, partial petalodics, and some doubles 



* For information on this point I am obliged to Mr Leonard Sutton 

 and to Mr Wootten. 



