I TIF. AMI'RK'AX IK^TAXIST 171 



differences to leiini, if [)()ssible, whether they were of any 

 special significance. On this basis, numerous old and familiar 

 species were reported U> be complexes horn which a number 

 of new forms could be derived. The result of such activities 

 on the science of botany was to t)bscure what before was 

 reasonably clear and to add to the perplexities of the begin- 

 ning student who was too often lo>t in a maze of species which 

 seemed all pretty much .dike. For some time thoughtful 

 students have felt thai a re-examination of the evidence upon 

 which such species were founded would be desiral)le, but the 

 first tangible results of investigations in this direction have 

 only recently appeared. Now the published results of such an 

 iiKpiiry into the species of Atriplcx, Chrysotliammis and Ar- 

 foiNsia by Frederic E. Clements and Dr. Harvey E. Hall 

 seem to prove the desirabilit}- of similar studies in other gen- 

 era. These investigators set out to examine, one by one, the 

 characters relict! upon to distinguish species in the genera 

 mentioned with a view to deciding what are and wh:it are not 

 of diagnostic inii)ortance and thus establish a scientific basis 

 for classification. Unlike other students they did not rely 

 solely upon herbarium material but studied the plants in the 

 field from Mexico to Canada and also made numerous experi- 

 ments with the plants under cultivation. Collectors of her- 

 barium six'ciniens are noted for selecting llic more striking 

 plants and in consequence such specimens are ciuite misleading 

 and not representative of the form as a whole. In the study 

 referred to it was found that many plants described as good 

 species were mere freaks or founded on characters that vary 

 greatly in all the species. The old way of making a species 

 was to label it new if it looked different; the new way endeav- 

 ors to examine the differences to see if they are stable and of 

 specific importance. In a survey as comprehensive as this it 



