58 THE PLANT SPECIES 



ning a specimen before identifying it. Bacteriologists, on the 

 other hand, are accustomed to carrying out physiological tests 

 with live specimens. In the case of sibling species in Drosophila 

 or Anopheles, etc., the dipterist may have to emulate the bac- 

 teriologist by employing physiological tests. 



In a similar way phanerogamic taxonomists usually identify 

 species by characters visible with the naked eye, a hand lens or 

 a stereoscopic microscope, whereas cryptogamic botanists have 

 long been habituated to the use of a compound microscope in 

 taxonomic work. The size of pollen grains and guard cells in flow- 

 ering plants is a microscopic character which varies proportion- 

 ately with the level of ploidy in some ( but not all ) polyploid com- 

 plexes. Where sibling species differ in ploidy level this character 

 may prove diagnostic. Preliminary studies carried out by the 

 author and an advanced botany class on herbarium material in 

 both the Artemisia tridentata and A. vulgaris groups have re- 

 vealed significant differences in pollen grain and stomatal size 

 in each one of these polyploid complexes, some of which can be 

 correlated with known chromosome numbers. Species recognition 

 in these and other critical groups might be facilitated by meas- 

 urements of cell size. Of course the chromosome number itself 

 is a morphological character. Nothing but convention stands in 

 the way of using characters which require a magnification greater 

 than 25 X or 50 X in the keys to difficult genera of higher plants. 



In the last analysis, if the search for workable macroscopic or 

 microscopic characters has been unsuccessful, it may prove neces- 

 sary to give up the attempt to distinguish the specimens of cer- 

 tain sibling species in the herbarium. There will be no objections 

 on the part of the biosystematists who make the segregations in 

 obscure groups to the filing of their indistinguishable species in 

 the same folder. The species still stand, however, as objectively 

 real biological units, whether the taxonomist can recognize them 

 on visible characters or not. A system of classification of the bio- 

 logical species must be judged, not on the basis of convenience, 

 but according to whether it represents accurately or inaccurately 

 the realities in nature. 



Gradual Formation of Species. Contrary to the belief held by 



