DEPARTMENT OF EXPERIMENTAL EVOLUTION. 8l 



De Vries has contended that varieties are characterized by the dropping 

 out of a unit character, and are hence retrogressive; that it is these cases 

 which show MendeUan inheritance. Progressive steps, he contends, take 

 place in several characters at once and are not inherited in a Mendelian way. 

 So many of the characters which give Mendelian inheritance can be explained 

 as an "unpacking" or dropping out that this position of de Vries has received 

 much favor. One who holds to this view is inclined to explain any apparent 

 progressive character which has Mendelian inheritance as being a reappear- 

 ance of an ancestral character, or, by a convenient logical device, the loss of 

 an inhibiting factor. I have been interested, therefore, in finding one charac- 

 ter having ^lendelian inheritance which seems to be incapable of being thus 

 explained away. 



The coccinellid pattern is made up of a few definitely placed spots or 

 coalescence of these spots into bands or vittse. Hippodamia extensa is one 

 of the few exceptions. This form must be considered a variety of H. conver- 

 gens, because it occupies a part of the range of that species and regularly 

 interbreeds with it. The extensa pattern is dominant to H. convergens, al- 

 though some heterozygous individuals show some dilution of the pattern. 

 In H. extensa a spot is replaced by a large and characteristic mark, covering 

 the position of the spot and considerable of the space in front of and to the 

 side of it. If a large number of specimens be examined, however, the extreme 

 fluctuants in the direction of reduction of pigm.ent show that this mark con- 

 sists essentially of a new crescent-shaped pigment center in front of the 

 spot referred to. I know of no spot in this position nor of this shape in all 

 the family, except that on one individual of H. glacialis. We must have, 

 therefore, in this mark a truly progressive Mendelizing characteristic. 



Certain environmental features are capable of modifying the parts of the 

 soma of some species when applied to the prepupa and pupa. The distinction 

 between the various factors which can or can not modify and also between 

 the parts which can or can not be modified, seems curiously arbitrary. The 

 solution must lie with the students of biochemistry and physiology. To the 

 evolutionist this susceptibility of a part of one species to an environmental 

 feature and the insusceptibility of the same part in another is an enlighten- 

 ing consideration. At least one of my experiments favors the view, for 

 which Tower's experiments give such strong evidence, that the germ and 

 soma are sometimes modified in a parallel way by an environment, so that the 

 result is an inheritance of the acquired character, which has arisen blastoge- 

 netically. H. convergens subjected to cold in the prepupal and pupal stages 

 has given a pronotum that is characteristic of the mountain species of this 

 genus, which are closely allied. Yet this feature is hereditary in the moun- 

 tain species when bred in our vivarium- Since it is of no utility, it doubt- 

 less arose in response to the environment in both germ and soma. 



Without entering into further details, the results obtained support the 

 following views with respect to the evolution of the color pattern in these 

 beetles at the present time: 



(i) Natural selection operating through a differential death rate, which 

 should be distinguished as lethal selection, is here of minimum im.portance. 



(2) Fecundal and sexual selection are also unimportant factors. 



(3) Yet evolution is very active. It takes place through determinate varia- 

 tion and differential potency. The steps are largely mutative, but in other 

 cases would seem to be flowing. The determinate variation, although funda- 

 mentally bathmic, is sometimes called out by environmental stimuli. 



6 — YB 



