SPECIES CROSSES IN THE GENTJS NICOTIANA. 223 



69/08 : Offspring of another similar plant of 70/07, which 

 differed from the last in having yellow flowers. Only 5 plants 

 raised, 3 yellow and two white, otherwise hke the parent. 



70/08 : Offspring of 71/07. See Tables IV. and V. 2 dwarf 

 plants (yellow bulged). 



71/08 : Offspring of 72/07. See Table V. The three f plants 

 include one of type E. All had white petals and white pollen. 



72/08 : Offspring of 72/07, another plant which Mke the 

 last had bulged flowers and white coroUa and pollen. Among 

 9 plants there were 2 of type E. 



v.— OTHER CROSSES. 



A considerable number of F 2 plants was raised from crosses 

 in which N. forgetiana was concerned as one of the parents. 

 Some of these crosses, especially those in which N. langsdorfii 

 took the place of a second grand-parent, appear hkely to give 

 rise to interesting novelties from a horticultural point of view. 

 The colour of the flowers in several of these crosses showed 

 signs of segregation into a redder and a bluer series, both of 

 which included several different shades. The colours seen 

 recall to some extent those shown by certain hybrid primulas. 

 The number of plants seen is not sufficient at present to 

 warrant any definite numerical conclusion, except that the red 

 series was more numerous than the blue. 



A single plant with varigated flowers arose from one of these 

 crosses, and its progeny will be watched with interest. 



A long series of coloured drawings of the flowers of these 

 crosses has been prepared by Mi*. A. de x41wis, Draughtsman to 

 the Royal Botanic Gardens, so that it should be possible to 

 make a close comparison with the characters of the next 

 generation. 



%.— CONCLUDING REMARKS. 



The experiments described in the preceding pages are only 

 a selection from a series of pioneering attempts to throw light 

 upon a branch of the science of heredity, to which statistical 

 observations have not hitherto been applied. One of the 

 principal object-s of these experiments was to ascertain how 



