542 



SCIENCE. 



[N. S. Vol. XXI. Xo. 536. 



From this last consideration we pass 

 naturally to a discussion of the nature of 

 the material which may be of use in the 

 study of fluctuating and discontinuous 

 variability. It needs no argument to sup- 

 port the assertion that a successful experi- 

 mental analysis of the behavior of separate 

 characters may be carried out only w^hen 

 dealing with series of organisms fluctuating 

 about a known mean with a measurable am- 

 plitude of variability. 



Systematic species as ordinarily accepted 

 generally consist of more than one inde- 

 pendent and constant subspecies, or ele- 

 mentary species which may not be assumed 

 to interbreed or intergrade, unless actually 

 demonstrated to do so by pedigreed cul- 

 tures. So far but few elementary species 

 have been found to interbreed. A due 

 recognition of this simple fact would save 

 us a vast amount of pyramidal logic resting 

 on an inverted apex of supposition. 



Again, the accumulation of observations 

 upon the prevalence and effect of self- and 

 cross-fertilization has totally unsettled the 

 generalizations current within the last few 

 decades. Briefly stated, a moderate pro- 

 portion of the flora of any region is 

 autogamous, a large proportion both 

 autogamous and heterogamous, and a mod- 

 erate proportion entirely heterogamous. 

 The relative number of species in- 

 cluded in the categories indicated varies 

 greatly in difiPerent regions. To assert the 

 deleterious effects of self-fertilization of 

 all, or a majority of plants, is to base a 

 statement upon evidence that lacks au- 

 thentication, as has been strikingly demon- 

 strated by recent results. As a matter of 

 fact no phase of evolutionary science is as 

 badly in need of investigation as that which 

 concerns the efTects of close- and cross- 

 breeding. 



It is also to be said that current miscon- 

 ceptions as to the extreme range of fluctu- 

 ating variability of many native species 



have arisen from a failure to recognize the 

 composite nature of the Linnean, or group- 

 species upon which observations have been 

 based, as I have found with the common 

 evening-primrose. 



The demands of ordinary floristie work 

 are usually met by the formulation of col- 

 lective species, which are an undeniable 

 convenience, and perhaps a necessity for 

 the elementary teacher and the amateur. 

 Upon the specialist in any subject rests 

 the obligation to furnish his non-tech- 

 nically trained constituency with concep- 

 tions of the facts and principles within 

 the domain of his investigations, which will 

 be inclusive, and easy of comprehension. 

 But if, in accordance with this requirement, 

 the systematist contents himself with this 

 looser, and with due regard it may be said, 

 more superficial treatment, and does not 

 delineate clearly the elementary constitu- 

 ents of a flora, or falters in carrying his 

 analysis of relationships to its logical end, 

 he fails notably in the more serious pur- 

 pose of his investigations, and his work 

 must be supplemented and extended before 

 it becomes an actual basic contribution to 

 the physiologic, or phylogenetic, branches 

 of the science. To study the behavior of 

 characters we must have them in their sim- 

 plest combinations. To investigate the 

 origin and activity of species we must have 

 them singly and uncomplicated. 



Lastly, we may turn to a phase of the 

 subject which has, as yet, received nothing 

 but speculative consideration — that of the 

 causes which induce the organization of 

 new characters and which stimulate their 

 external appearance. The recurrence of 

 the known mutants of Lamarck's evening- 

 primrose and the occurrence of new mu- 

 tants of other species have taken place in 

 New York and Amsterdam under condi- 

 tions that lead to the definite conclusion 

 that a favorable environment, including 

 the most advantageous conditions for 



