RELAPSE AND RETROGRESSION. 81 



hand, the yellow of the allied Lysimachias certainly 

 appears to be primitive. 



Among heaths, our only yellowish sort is the 

 very degraded Monotropa, a leafless saprophyte of 

 the lowest type, obviously a product of extreme 

 Retrogression. 



The colours of many Scrophularine<z may be ex- 

 plained in the same way. Perhaps the yellow of the 

 mulleins is primitive ; but as some species are white or 

 purple, it is just as likely to be retrogressive. In 

 Linaria, we may almost be sure that retrogression has 

 taken place ; for we can trace a regular gradation from 

 lilac flowers with a yellow palate, like L. cymbalaria, 

 to pale yellow flowers, like L. vulgaris (Fig. 25), which 

 has the mass of the corolla primrose, and the palate 

 orange. Mimulus luteus is also yellow, but it is usually 

 marked inside with small purple spots, and sometimes 

 has a large pink or red patch upon each lobe. In 

 Melampyrum cristatum, the yellow corolla is variegated 

 with purple : in M. pratense, it has the lip deeper in 

 hue. All these genera include many purple and 

 variegated species ; and the yellow members almost 

 always bear evident marks of being descended from 

 polychromatic ancestors. 



Among Lentibulacea, our debased Utricularice have 

 pale yellow flowers ; but Pinguicula vulgaris is bluish 

 purple, and Pinguicula lusitanica yellow, tinged with 

 lilac : so that here, too, we may suspect Retrogression. 



The case of the yellow Composites, especially the 

 Ligulatce, is more difficult to decide. It would seem 

 as though these plants, which have all their florets ligu- 

 late (Fig. 27), must be more highly developed than the 

 Corymbiferce, which have only the ray-florets ligulate, 



G 



