MISCELLANEO US. 113 



calyx bright red, instead of by increasing the size 

 and deepening the hue of its small white petals. In 

 the Fuchsia, the Hydrangea, and many other well- 

 known exotics, we get exactly similar devices. 



Parasites and saprophytes do not as a rule require 

 to produce green leaves ; hence, most of them, like 

 Cuscuta, Orobanche, Lathrcsa, and Monotropa, have 

 the stem and leaves (or scales) coloured like the 

 flowers. Imperfect parasites which contain chloro- 

 phyll, however, have the leaves more or less green, 

 as in Viscum, Bartsia, RJiinantJiiis, and Melampyrum. 



The outer florets of compound heads are apt to 

 produce larger petals than the inner ones, as in many 

 Umbellates (like cow-parsnip), the guelder-rose, the 

 Hydrangea, and the rayed forms of Composites. These 

 are obviously intended to increase the total attractive 

 effect. In the Umbellates and in candytuft the outer 

 petals of the individual flowers grow longer than the 

 inner ones. 



Petals have perhaps been independently developed 

 from .stamens at least twice over, once in the Dico- 

 tyledons, and once in the Monocotyledons. Insects, 

 having once learnt to visit coloured surfaces in search 

 of pollen and honey, would naturally tend to visit all 

 such surfaces in future, and thus to select for fertili- 

 sation any coloured flowers that offered them any 

 attraction in the way of food, of whatsoever sort. 

 Apparently, at last one species of gymnosperm, the 

 larch {Pinus larix), has thus become entomophilous, 

 its fertile scales being interspersed with bright pink 

 or red empty bracts, which seem to subserve an 

 attractive function. They are certainly visited by 

 insects, perhaps in search of some secretion from 



I 



