101 



}-//>/vyx//x. etc. These are fortuitous variations, for I cannot 

 believe that the disposition of these marks where, as in these 

 cases, they take every form that is conceivable, can be of any 

 benefit to the species, any more than the mere variation in the 

 number of lobes in the leaves of different oaks growing under 

 like conditions can be of any particular benefit to the species, 

 however useful to us in classification. 



In my address before the Section of Biology of the American 

 Association for the Advancement of Science, at Cleveland, in 

 1888, 1 have discussed the various forms and causes of variation, 

 and especially the limitations of natural selection, stating ex- 

 pressly that this last " deals only with variations useful to the 

 organism in its struggle for existence, and can exert no power 

 in fixing the endless number of what, from present knowledge, 

 we are obliged to consider fortuitous characters," and I have 

 long recognized, from my studies of insect life, the existence of 

 these fortuitous variations. The subject has since been very 

 well elaborated by Professor Ward in his communication to 

 the Society (December 15, 1888) on k ' Fortuitous variation as 

 illustrated by the genus Eupatorium " and in his Annual 

 Address (January 24, 1891) on u Neo-Darwinism and Neo- 

 Lamarckism," and the Prodoxida? furnish an excellent illustra- 

 tion of this fortuitous variation. Yet at the same time that we 

 note this chance variation, as exemplified in a number of the 

 species of Prodoxus, which are mere ravagers or despoilers and 

 have not been brought into any special or mutual relations with 

 the plant, we have, on the other hand, in Pronuba yuccaseila, cor- 

 related with the other striking structural modifications which 

 have brought it into such special relations with the plant, an 

 elimination of all maculation or markings upon the primaries, 

 and a purely white coloring so fixed that it shows absolutely no 

 variation over half the continent. The structural variation has 

 boon necessary a consequence of effort, environment, and nat- 

 ural selection. The color variation, on the contrary, has not 

 been absolutely necessary, yet has nevertheless gone on in lines 

 which, tending to give greater protective resemblance to the 

 flower, have in the long run proved to be, perhaps, the most 

 advantageous. I thus recognize three distinct lines of variation 

 as exemplified in these Prodoxida^, and what is true of them is, 

 I believe, true of all alliances of organisms. The first and most 

 important is structural and generic ; it is absolutely essential 

 and is preserved in its perfection by the elimination, through 



