359 



A PRELIMINARY ACCOUNT OF AN INTERSPECIFIC 

 HYBRID AND BACKCROSSES OF DIGIT A J.I S. 



BY 



Ernest Warren, D.Sc, 

 Director of the Natal Museum, Pietermaritzburg. 



With Plates IV, V. 



Read July 12, 1921. 



For a number of years I have made a study of the inheritance 

 of certain characters in the ordinary garden foxglove {Digitalis 

 gloxiniceflora) , and some of the results have been published in 

 "Biometrika" 1 and in the "Journal of Science." 2 



I have also made certain observations on the hybrid between 

 Digitalis gloxiniaflora and Digitalis hi tea and on the backcrosses 

 of the hybrid with the parent species; and it is proposed to give 

 a brief summary of the investigation. 



It is presumed that D. gloxiniajlora is simply a cultivated 

 variety of D. purpurea. In any case it is quite fertile and breeds 

 perfectly true on «elf -fertilisation, and from a genetic standpoint 

 it acts as a pure species. 



1. Gloxiniceflora (Q) x lutea (d*)- 



Twenty-five plants of gloxinia flora with varying characters 

 as to shape of flower, colour, etc., were pollinated from six plants 

 of lutea. Five flowers in each plant were fertilised. Out of these 

 125 ovaries, 45 shrivelled very early without result; the remaining 

 80 grew to a varying size and ripened slowly. Of the ripe ovaries 

 12 were small and contained nothing, the rest (68) bore some seed. 

 The seeds were very markedly dimorphic, a few were large and 

 broad, while the great majority were very small and relatively 

 narrow (PI. IV, fig. 1). On the average each of these capsules 

 bore only 4 large seeds and some 2-300 minute seeds. Even the 

 large seeds differed conspicuously from the seeds obtained by the 

 self-fertilisation of the gloxinias flora parent. 



The hybrid large seeds were of about the same breadth as that 

 of the ordinary seeds of the female parent but they were very con- 

 stantly shorter, and consequently appeared squat or squarish. This 

 was not due to the seed accommodating itself to an embryo inter- 

 mediate in shape between that of gloxi nice flora and hi tea, since 

 although the seed of lutea is larger than that of gloxiniceflora, yet 

 the average ratio of length and breadth is practically the same 



1 Warren, E. " Breeding Experiments with Foxgloves," Biometrika, Vol. 



XI, 1917. 



2 Warren, E. "The Pure Line Hypothesis etc.," S.A. Journal of Science, 



Vol. XV, 1918-9. 



