OF WASHINGTON. 17 



constrained to admit that their favorite principles must have raw 

 materials to work upon. In other words, the advantageous 

 structure or habit must reach the point of being advantageous 

 without the assistance of the principle which then stands ready 

 to encourage its development. Notwithstanding this defect, 

 the theory of selection as variously analyzed and sub-divided has 

 been and is still advanced as an adequate explanation of the 

 modus opcrandi, if not as the actual cause, of evolution. 



In groups where complex adaptations to external environment 

 have taken place, the issues are so mixed that contributions of 

 natural, sexual, germinal, or other forms of selection and their 

 resulting coordinations can scarcely be estimated, even in the 

 most general way. The nicety of many adaptations has so en 

 couraged the imaginative mind that extended flights of fancy 

 have not unfrequently passed as sober theory, if not demonstrated 

 fact. Of a bird or an insect blown to a new region, where it 

 changes its climate, food, habits, and even its enemies, much 

 may be predicated with comparative safety, and with sufficiently 

 numerous factors of undetermined importance supplied by con 

 veniently fortuitous circumstances, selection may be made to ap 

 pear as the main-spring as well as the balance-wheel of creation. 

 But among the diplopods, at least, the simplicity and uniformity 

 of external adaptation, considered in connection with the quan 

 tity and constancy of the internal diversity of the group, seems 

 to warrant a view for which analogy indicates a wider applica 

 tion. Diversity is seen to be essentially independent of selec 

 tion. Selection accelerates, retards, or even reverses evolution, 

 but by interference in a succession of changes which it does not 

 cause. Segregation, of whatever kind, permits the accumula 

 tion of variations which it does not initiate nor direct. The bi 

 ology of the Diplopoda indicates that facilities in the way of 

 segregation have long been ample for the differentiation, not 

 alone of species and genera, but of families and orders. To 

 prove with reference to any individual case the negative propo 

 sition that no selection has intervened would be impossible, but 

 the honest student will find the Diplopoda and other groups of 

 similar ecologic status replete with instances of which the 

 unique dorsal processes of the OxydesmidaB are a minor example. 

 As a single more general illustration may be mentioned that of 

 the repugnatorial pores, constantly present in some orders and 

 constantly absent in others, but in the Merocheta having the 

 peculiarity of occurring in interrupted series. The presence, or 

 rather the absence, of these organs may well be correlated with 

 habits or environment, but that the omission of the pores of a 

 single segment or two segments from an otherwise continuous 

 series is a matter of selective advantage or disadvantage is very 

 difficult to believe. 



