OCTOBEB 14, 1910] 



SCIENCE 



519 



issued, 160 pp. Most of the articles treat of 

 insects injurious to man or animals. 



Some years ago Dr. 0. M. Eeuter published 

 a system of classification of the hemipterous 

 family Capsidae. Now he has issued a new 

 arrangement.' He has modified his previous 

 classification in various details and made nine 

 subfamilies. He gives a list of the genera, 

 placing most of them in the proper subfamily. 

 The article also includes a review of the classi- 

 fications of the Heteroptera, and a new one, 

 in which he arranges the 40 families in 12 

 superfamilies. There are tables to these fam- 

 ilies and to the groups of the Capsidse. One 

 of the new features is the elevation of Piesma 

 to family rank. 



Nathan Banks 



sj 



SPECIAL ARTICLES 



THE SELECTIVE ELIMINATION OF ORGANS 



One of the monuments erected to Charles 

 Darwin on the hundredth anniversary of his 

 birth might have been a bibliographic index 

 to the literature of organic evolution. But it 

 is very much easier to pen a series of ad- 

 dresses on Darwin's method, Darwin's real 

 opinion, Darwin's influence, than it is to com- 

 pile a comprehensive bibliography and an- 

 alyze it with the thoroughness and detail and 

 wisdom necessary to make it a really useful 

 aid to the investigator; it would have taken 

 a very plucky librarian (with wealthy friends 

 and a genius for interesting them in his 

 undertakings) to carry it through. 



As his card manuscript for the subject in- 

 dex approached completion he would have 

 found that several drawers in his cabinet 

 were required for the cards bearing the cap- 

 tion natural selection. These cards would 

 have been a key to everything that can be 

 said in a theoretical way about natural se- 

 lection. The student who would take these 

 cards and attempt conscientiously to cover 

 the field would be ready, after a year's floun- 

 dering about in the morass of rhetoric, to be- 



' " Neue Beitriige zur Phylogenie und System- 

 atik der iliriden," Ada Soc. Sd. Fenn., XXXVII., 

 No. 3, 1910, pp. 171. 



lieve that all the arguments — for and against 

 — have been presented in all their possible 

 permutations. 



That no solid foundation for a scientific 

 superstructure is to be found in this polemic 

 quagmire has often been recognized; at pres- 

 ent natural selection is out of fashion among 

 biologists. Other problems are in the search- 

 light. 



It is quite natural that a theory which has 

 been so much talked about but as little in- 

 vestigated should cease to be attractive at a 

 time when concrete experimental proof is so 

 much in demand. But can not such proof be 

 adduced for natural selection? Is it not pos- 

 sible that the biologist of to-day with the 

 powerful tools of statistical analysis at his 

 service may be able to demonstrate the exist- 

 ence of natural selection, just as by the use 

 of these tools he has been able to measure the 

 strength of heredity? 



Fortunately a beginning has already been 

 made, for if the index were brought well up 

 to date probably over a dozen of the cards in 

 the drawers devoted to natural selection would 

 bear titles of papers embodying the results 

 of serious attempts to measure the intensity 

 of the selective death rate in some organism. 



In the selection theory of evolution — the 

 pure Darwinian theory as popularly con- 

 ceived — there are three factors which must 

 be not only existent, but coexistent, if there is 

 to be any shift in the characteristics of suc- 

 ceeding generations of any organism. These 

 factors are variation, inheritance and selective 

 elimination. If any one of these be absent or 

 its force counterbalanced by some other fac- 

 tor. Darwinian evolution in that species can 

 not be taking place at the moment in ques- 

 tion. 



Now a great mistake of most of the men 

 who have written on organic evolution has 

 been that they have tried to solve the whole 

 problem. Lacking data (or having only a 

 modicum of data), they have invoked as- 

 sumptions and logic, and, having proved their 

 assumptions by their logic, have proceeded to 

 generalizations. In dealing with a problem 



