io8 



THE COLOURS OF FLOWERS. 



involucres, bracts, glumes, glands, awns, and so forth. 

 Now, to which of these classes do the yellow flowers 

 ordinarily belong ? Clearly to the first. To which 

 do the green flowers ordinarily belong } Clearly to 

 the second. The organisation of the catkins, the 

 sedges, and the grasses is exactly analogous to that 

 of the spurges, which we know by an unbroken line 

 of intermediate links to be descended from petali- 

 ferous ancestors. The inference is almost irresistible 



a.i. 



Fig. 44 —Diagram of flower of grass, a, sepals ; a i, outer sepal, flowering glume. 

 or outer palea ; a 2 and a 3. inner sepals, combined into a single inner palea ; 

 b, petals ; b 1 and /5 2, the lodicuks ; b 3, suppressed ; c, stamens, all present \ d, 

 styles or stigmas; d\ and ^2, present ; ^3, suppressed. The whole inner side 

 of the flower is thus abortive. 



that so highly complicated a flower as that of the 

 grasses, with its one-celled, one-ovuled ovary, its two 

 styles, and its advanced paraphernalia of lodicules, 

 paleae, and glumes, arranged in long and subdivided 

 spikes, must be a very specialised or degenerate, not 

 a primitive or early type. The more closely we 

 examine green flowers, the more do we see that 

 they form the opposite pole to such simple and truly 



