456 KINSHIP AND ADAPTATION 



analog}' with natural selection as Darwinians commonly 

 assume. Nevertheless, cautious observation of plants and 

 animals under domestication is sure to throw important 

 light upon what happens in nature, for the artificial conditions 

 being more under control, it is easier to estimate the effects 

 which a given factor (such as heredity for instance) may 

 safely be counted upon to exert. Some quite unexpected 

 results have been reached through studies in the history of 

 garden vegetaljles by Dr. E. L. Sturtevant of the New York 

 State Agricultural Experiment Station. Wishing to gain 

 what evidence he could of "the extent of variation that has 

 been produced in plants through cultivation," he examined 

 all pictures and descriptions in the old herbals — many of 

 them entirely trustworthy records although published two 

 or three centuries ago — and compared them with the more 

 important recent records and with living examples of the 

 forms now in cultivation. It seemed not unreasonable to 

 expect that among so many plants, cultivated for centuries, 

 at least a few examples would be found in' which extreme 

 types such as cauliflower, Brussels s])routs, kohlral)i, and 

 other derivations of the wild cabl)age, might be connected 

 y>y a series of slightly differing connecting links. In not a 

 single case has such a series been found. So far as may he 

 judged from the evidence>, new types are not developed from 

 other types by the cumulative seliH'tion of slight variations 

 in a given direction; ])ut the}' come Kuddenl}^, and each is 

 distinct from its first appearance. All that the selection of 

 slight variations ever accomplishes is the improvement up 

 to a certain ])oint of features already well marked; and ex- 

 perience shows that tliis point is soon reached. A comjiact 

 fleshy inflorescence, as of the cauliflower for instance, may 

 be made more com])act and more fleshy within certain rather 

 narrow limits, and the highest degree of perfection can be 

 maintained oul}' by the most careful cultivation and generous 

 enrichment of the soil. \\'e lia\'e already seen that carrots 

 under (cultivation exhibit similar limitations. Such facts 

 are as unfavorable to Lamarckism as to Darwinism since 

 both su|)])ose that types hav(> always evolved through slight 

 changes. 



