June 7, 1907] 



SCIENCE 



885 



gions, could produce a type varying about 

 a new average, which lies near one of the 

 extremes of the fluctuating variations of 

 the parent species. 



If this new average should be estab- 

 lished within the limits of the transgression 

 variations of two species one of which 

 existed in northern Europe and the other 

 in central Europe, we should have the 

 production of similar types, the alpine and 

 arctic type, in widely separated regions 

 and from different parent species. The 

 characters of the new type are not 'fixed' 

 in the sense of being due to inheritance, 

 but only in the sense that they are a re- 

 sponse to a particular combination of ex- 

 ternal factors, and this combination is 

 constant in the given regions. Such a 

 view of the origin of alpine types is not 

 merely of theoretical interest, since the ap- 

 plication of the physiological method gives 

 the means of reaching more or less definite 

 conclusions. 



De Vries and others have pointed out 

 that the species of the manuals and the 

 systematic botanists are in large part com- 

 posite or collective species and not simple 

 or elementary species. In his view the 

 latter differ from their parent species by 

 new characters, not by modifications of old 

 ones. The new characters are inheritable 

 as soon as they appear, and are not regu- 

 lated by the external conditions in which 

 the adult plant lives. 



If MacDougal's work stands the test of 

 repetition, physiological experiment may 

 open up a new field in investigating the 

 origin of species. One method of apply- 

 ing physiological experiment to determin- 

 ing the limits of species has just been dis- 

 cussed. But other applications of this 

 method are possible. It is well known that 

 cross fertilization generally takes place 

 only between closely related species of 

 plants, rarely between genera. "When at- 

 tempts are made to cross species remotely 



related, either the pollen does not grow 

 upon the stigina of the strange species, or 

 fertilization of the egg does not take place, 

 or if seeds develop the resulting hybrid is 

 sterile, not being able to produce seeds for 

 its propagation. What lies at the basis of 

 these physiological differences is still ob- 

 scure. It is probable that enzymes, toxins 

 or other chemical substances play a part. 

 But whatever the explanation, the fact 

 may be used in determining the nearness 

 or remoteness of the relationship between 

 forms. This possibility has been recog- 

 nized by many investigators, and biologists 

 have proposed using the degree of fertility 

 of hybrids as the means of distinguishing 

 genera, species, and varieties. Though this 

 has been found not to be reliable in all 

 eases, de Vries has suggested it as a means 

 of distinguishing his elementary species 

 from varieties. If on crossing two forms 

 the resulting hybrid is constant in regard 

 to a given character, when guarded against 

 further crossing, the two forms were dif- 

 ferent species. But if, on crossing, the 

 descendants of the resulting hybrid fol- 

 lowed Mendel's law of hybrids, according 

 to which one fourth of the offspring of the 

 hybrid in each succeeding generation re- 

 sembled one parent in respect to a given 

 character, one fourth resembles the other 

 parent as regards the corresponding char- 

 acter, while half are like the original hy- 

 brid, then the parent forms of the hybrid 

 were one and the same species. 



Whatever the limitations of this method 

 in its practical application, the significant 

 fact is the extent to which physiological 

 conceptions have invaded a realm that was 

 purely morphology. We may use the ex- 

 perimental method in studying the origin 

 of new species and varieties. We may 

 apply physiological methods in determin- 

 ing the range of the fluctuating variations 

 within the species. We may use physi- 

 ological affinities as the test of the degree 



