i6 
James Small. 
his earlier papers (58) may be given as a diagram, Fig. 8. In his 
Alpenblumen (59), Table II, p. 481, bis figures for the insect visitors 
to yellow and to white flowers are very similar, but he gives the 
colour sequence—greenish-yellow, yellow, yellow with orange red 
spots, white, white with yellow spots, rose red, deep red. From 
other tables butterflies and even bees seem to prefer yellow to 
white. 
Avebury (2) considers that blue flowers have evolved from 
green flowers by passing through the intermediate stages of white or 
yellow and generally red. 
Grant Allen (1) gives a similar colour sequence, but in 
considering the Compositae he takes the Cynarese as primitive on 
account of their highly developed colours and states that reversion 
in colour has been the rule throughout the family because “ the 
primitive ancestral composite had reached the stage of blue or 
purple flowers while it was still at a level of development 
corresponding to that of the scabious or jfasione." He also 
considered that the “ Ligulates were again developed from the 
yellow-rayed Corymbifers by the conversion of all the disk florets 
into rays.” 
Observations by Willis and Burkill (92) confirm the general 
sequence. Table XII (92a) shows yellow above white but Table 
XIII ( loc . cit .) shows a decided preference by the higher insects for 
white. 1 In most of the tables given white is seen to be preferred to 
yellow by desirable visitors and in Table XXXVII (92b) the order 
of efficiency is given as—rose purple, blue, lilac, white, yellow, 
green. 
De Vries (87-88) gives many examples in the Compositae of the 
atavistic variations noted by Hildebrand (37) and also progressive 
variations where white flowered species have suddenly developed blue 
varieties, such as Ageratum coeruleuin. A similar progressive 
mutation, where a white ray appeared in one capitulum of a normally 
yellow flowered Dahlia, is discussed in its biochemical aspect by 
Kajanus (45). Other sudden variations in colour are recorded in 
Cosmos by Longo (50). An interesting atavism is recorded by 
Daniel (23), see also Small (77). 
Anatomy. Jacquin (43) and other early botanists who figured 
the corolla in the Compositae did not usually indicate the venation 
accurately; usually a vein is put in the position of the midrib for 
1 There are several kinds of whites, see Neilson Jones (64a). 
