September 27, 1901.] 



SCIENCE 



469 



siasm in publication far outstrips their gen- 

 eral scholarship ; but one may express gen- 

 uine surprise that the heads of important 

 botanical departments and editors of promi- 

 nent journals let these nomenclatorial sole- 

 cisms see light in print. Here is another 

 opportunity for easy improvement in the 

 methods of systematic botany. 



To this point I have dealt chiefly with 

 the form of presentation. Let us now con- 

 sider the subject matter. Here the difficul- 

 ties of improvement are naturally greater. 



The first feature of this subject which 

 demands attention is the artificiality which 

 still lingers in our so-called natural system. 

 It is true that the natural arrangement of 

 orders and families has been much im- 

 proved in recent years. The clues derived 

 from the varying degrees of adnation, con- 

 nation and zygomorphy of floral parts in 

 the dicotyledons have suggested the first 

 system in which groups of such obvious 

 affinity as the Caryophyllaceae, Aizoaceae, 

 Scleranthaceae and Amaranthaceae are found 

 in natural proximity. But much artifici- 

 ality still remains in the details of modern 

 classification. For instance, we are com- 

 monly treating as equivalents in our system 

 things which in nature have widely dif- 

 ferent values. 



There is an old question always coming 

 up, ever fresh for discussion, never very 

 clearly settled, regarding the objectivity of 

 species. Do they exist in nature or are 

 they artificial categories ? Much may be 

 said on both sides. It takes, however, no 

 very profound study of plants and their 

 descriptions to reveal the fact that so-called 

 species are of both kinds. Many thousands 

 exist as well-marked entities in nature, but, 

 alas ! there are many hundreds more which 

 scarcely extend beyond the subjective. 

 They represent not permanent lines of more 

 or less independent development in nature, 

 but chance combinations of inconstant 

 characters analogous to cross-sections 



through some plastic and still unsolidified 

 material. 



The cause of this lies partly in the author 

 of the species and is partly inherent in na- 

 ture. On the one hand, such so-called spe- 

 cies may result from the hasty description of 

 plants whose differences, observed in a few 

 herbarium specimens, have not been suffi- 

 ciently verified in the field. On the other 

 hand, they may come from the simple fact 

 that there are no formed or settled species 

 in the group concerned. The forms of that 

 particular affinity are still in a state of free 

 intergradation and the species im Werden he- 

 griff en. 



There seems to be a wish upon the part 

 of many systematists to ignore this fact ; to 

 maintain that this or that form is, in hack- 

 neyed phrase, a ' perfectly good species ' 

 because it shows certain differences from its 

 slightly removed although copiously inter- 

 grading neighbors ; in fact, to asseverate that 

 all plants which show differences worthy 

 of remark should, irrespective of their con- 

 stancy, be classed as species. But not- 

 withstanding these unhappy ideas, nothing 

 can be more certain than that fortuitous 

 cross-sections in the nebulous places of na- 

 ture are not species in the sense that Ranun- 

 culus pennsylvanicus, Juncus trifidus, Malva 

 rotundifolia or Potentilla tridentata are. Nor 

 can we hope to escape great artificiality in 

 aqy system which assigns to like rank and 

 groups in the same category things of such 

 diverse nature and significance. 



Species as now recognized are not equiv- 

 alent things. The category, called specific, 

 is itself a complex, in the same need of crit- 

 ical studj', of subdivision, of segregation, as 

 manj^ of its elements. There are species 

 marked by pronounced morphological fea- 

 tures, which they never lose and which may 

 always serve to identify them. There are 

 others with characters subject to concom- 

 itant variations, in which if one feature 

 varies in a particular direction the change 



