122 Flowers and their Pedigrees. 



often of considerable size, and of very different habits 

 from our little goose-grass. But in the temperate 

 regions, since the earth has begun to cool into zones, 

 some of these rubiaceous plants have found out that 

 they could get along better by becoming little creep- 

 ing weeds ; and these are the stellates, including our 

 present friend. Accordingly they have mostly given 

 up the attempt to attract big honey-sucking insects 

 whose long proboscis can probe the recesses of jasmine 

 or woodbine, and have laid themselves out to please 

 the small flies and miscellaneous little beetles, which 

 serve almost equally well to carry their pollen from 

 head to head. Now the flowers which specially cater 

 for such minor insects are usually quite flat, so that 

 every kind alike can get at the honey or the pollen ; 

 and that, I fancy, is why the goose-grass and so many 

 of its allies have lost their tubes. They are, in fact, 

 somewhat degenerate forms, descended from highly 

 adapted tr(^ical types, but now readjusted to a 

 humbler though more successful grade of existence. 



Closely connected with this question is the other 

 and very interesting problem of their colour. Why 

 is goose-grass white ? For the very same reason — 

 because it wishes to attract all sorts of little insects 

 impartially. For this purpose white is the best 

 colour. Almost all flowers which thus depend for 



