470 



SCIENCE. 



[N. S. Vol. XIV. No. 352. 



is regularly accompanied by certain other 

 modifications affecting other members. 

 Such species may be subdivided, and have 

 good subspecies or varieties. On the other 

 hand, there is a totally different type of 

 species in which variation is not concom- 

 itant, in which one feature changes with- 

 out apparent connection with any other, in 

 which, for instance, thorns may be devel- 

 oped or be absent, while leaflets may be few 

 or many irrespective of the presence or ab- 

 sence of thorns, and again, the inflorescence 

 may show further variation quite independ- 

 ent of leaflets and thorns. Such species, ex- 

 hibiting what Dr. Gray called promiscuous 

 variation, are well illustrated by Acacia fili- 

 cina, Mimosa asperata, certain Aquilegias, 

 Delphiniums and Lupines. In these cases 

 segregation or even varietal subdivision, 

 although often attempted, has little or no 

 significance, for the segregates exhibit only 

 kaleidoscopic combinations of ever-chang- 

 ing characters. There are, on the other 

 hand, especially as the result of preponder- 

 ating close fertilization or vegetative repro- 

 duction, species which exhibit a wonderful 

 constancy of small characters, a remarkable 

 fidelity in transmitting from one generation 

 to the next the most obscure traits. Such 

 are the segregates of Draba verna, elaborated 

 by Jordan, studied with such keen interest 

 by de Bary during the last months of his 

 life and critically reviewed in the later 

 work of Rosen. Such are also the newly 

 recognized Alchemillas of the Alps and our 

 own Antennarias. 



From these illustrations it is easy to see 

 that species as now recorded in literature 

 are by no means alike and that they can- 

 not be regarded as equivalents in any com- 

 plete or logical system of classification. 

 Curiously enough, however, the term 

 ' species ' seems to be growing more and 

 more popular as it means less and less. 

 Often and on all sides we hear lengthy 

 arguments, and emphatic asseverations to 



the efiect that this or that plant is a ' per- 

 fectly good species ' ; and if in the course of 

 monographic work a so-called species is let 

 down to varietal rank it rarely fails to find 

 somewhere its ardent defenders, who ap- 

 pear to hold the curious view that the 

 monographer has not merely expressed a 

 scientific opinion, but has somehow perpe- 

 trated an injustice upon the plant or its de- 

 scriber. How anxious most discoverers of 

 new forms are that their plants may prove 

 species, not mere varieties, and finally what 

 a fascination the mere binomial appears to 

 exert upon certain minds ! Is it any won- 

 der under these circumstances that the spe- 

 cific category has been overcrowded and 

 made to include such widely different ele- 

 ments that the word species has lost nearly 

 all its taxonomic significance ? 



However, no thoughtful botanist who can 

 rise above a merely subjective attitude to- 

 ward the few species in which he chances 

 to' be particularly interested and take a 

 broader, more objective survey of the whole 

 field, will be satisfied that the present hodge- 

 podge of non-equivalent forms in the spe- 

 cific category represents the finished result 

 of a natural system of classification. 



Species must be subjected to a gradual 

 reclassification along more definite lines. 

 Overwhelming as the task may at first ap- 

 pear, it is fortunately one which can be 

 taken up little by little, a work in which 

 every systematist, every collector, every 

 amateur, who will, may take part. The 

 first step is evident enough. Each species 

 must be examined in the light of vastly 

 more copious material than at present 

 exists even in our largest herbaria. Has 

 there ever been a conscientious monog- 

 rapher who has not seen the pressing need 

 of further material in his group, who has 

 not felt that ten or even a hundred times 

 as many specimens would have been neces- 

 sary to yield a satisfactory knowledge of 

 the directions and limits of variation ? Let 



