i68 
Edith R. Saunders. 
the type is the more common form in the Alps, the variety in the 
Pyrenees, the variety being chiefly found on granite in Switzerland 
according to Schinz and Keller. 1 In Tirol, Vorarlberg and 
Liechtenstein, on the other hand, Dalla Torre and Sarntheiir state 
that the concolor form is almost more frequent than the type 
Certain stations are also mentioned by these latter authors when 
rosea —a much rarer form—has been observed to occur wild. 
Cross-breeding experiments were begun in 1898 and repeated 
in 1899, but the small number of germinations obtained and pressure 
of other work led to the abandonment of the experiments for the 
time being. More recently the experiments have again been 
repeated, and as the later results are in agreement with those 
obtained earlier, we may regard them as sufficiently conclusive, 
although the total number of plants raised is small. The number 
of germinations in each case was disappointingly few, but whether 
owing to failure to achieve the right conditions, or to the fact that 
much of the seed was not really good, it is difficult to say. Another 
circumstance also tended to limit the number of plants raised. It 
was found that an individual, if covered, or if isolated uncovered, 
generally failed to set seed; hence, in order to obtain the F 2 
generation without artificial fertilisation, it was necessary to place 
the F x plants at least two together under conditions in which they 
might be visited by visits without risk of being crossed with the 
parent forms. 
Results of Experiments. 
Three forms were employed, viz :— 
1, alpina type; corolla blue with orange palate. 
2, var. rosea ; corolla pink „ „ ,, 
3, var. concolor; corolla blue, orange absent from the 
palate which may be whitish or more or less deeply 
overspread with blue. 
A few crossbreds were raised in 1899 and 1900 from reciprocal 
matings between the type and concolor , and later others were 
obtained from matings between concolor and both the forms with 
orange palate. All the F x plants resembled concolor, the orange 
colour being absent from the palate. In some flowers the corres¬ 
ponding or a somewhat smaller area of the under lip was whitish in 
colour, in others on the same individual the whole palate might be 
1 Flora dcr Schweiz, 1900, p. 453. 
3 Die Farn- und Bliithenpflanzen von Tirol, Vorarlberg und 
Liechtenstein III, p. 250. 
