API!. 1896.] BOTANICAL SOCIETY OF EDINBURGH. 545 



indican and ruberythiic acid give rise respectively to 

 indigo blue and alizarin, is paralleled by the process by 

 which within the organism another glucoside (?) gives 

 rise to the colour of the rose. These facts are of more 

 importance than is apparent at first sight. They surely 

 suggest that whatever part the selection of insects has 

 played in the final development of the colour of the rose, 

 this is primarily due to the chemical properties of the 

 tannin which ontogenetically precedes it. Now tannin is 

 almost certainly a waste product, a result of destructive 

 metabolism. As such it probably tends to occur wherever 

 assimilation is feeble or suppressed. What the conditions 

 are which determine the conversion of the tannin into the 

 chromogen of anthocyan, rather than into the chromogen 

 of other pigments, such as the phlobaphenes, is of course 

 unknown, but it is probable that they are purely chemical. 

 It may of course be said that this is the point at which 

 selection intervenes, favouring the production of the useful 

 (i.e. attractive) anthocyans rather than of the useless dull- 

 coloured phlobaphenes. This, however, is doubtful on the 

 following grounds. When tissues are injured as by gall- 

 producing insects, pigment is usually produced, but this 

 may be either one of the phlobaphenes or anthocyan. As 

 it is difficult to see how such colouration can be subject 

 to selection, this furnishes a strong presumption that the 

 particular pigment produced is determined primarily by 

 chemical agencies. Again such a vigorous selection as the 

 statement would imply does not seem to be borne out by 

 facts. Thus, observations on our fruit-eating birds do not 

 lead one to suppose that the result of their preference 

 would be to favour the dissemination of red varieties of 

 apples rather than of the " russet " ones, supposing the size 

 of the apples to be such that they could be readily carried 

 off. In this case, the fact that " russet " apples, when cut, 

 become discoloured far more rapidly than red ones, suggests 

 that the russet colour is the consequence of the marked 

 tendency of the tannins of the apple to become converted 

 into phlobaphenes, and is, therefore, simply a consequence 

 of the chemical composition of the cell contents. 



Although so little is known as to the conditions upon 

 which the presence of anthocyan is dependent, yet the 



