206 DISCOVERY REPORTS 



GENERAL NOTES 



THE SPECIES AS A POLYPHASIC SYSTEM 



When attempting systematic work of any kind, sooner or later, the " species problem" 

 has to be faced. The problem in its widest aspect presents a twofold difficulty. In the 

 first place one has to frame a species concept, and, secondly, give it adequate termino- 

 logical expression. 



I conceive the species as an orbital system of disparate units expressing itself in time 

 and space, somewhat analogous to a solar system. Under normal conditions the units 

 are concentrated into genie clusters and scatter centrifugally as conditions become 

 adverse. The whole system is analogous to a complex kinetic system, in which the units 

 act and are acted upon at random. An association of several such discrete systems con- 

 stitutes a generic cluster. A collection of individual variants under any given set of 

 environmental factors constitutes what I call a phase. The conception of a species as a 

 polyphasic orbital system of genie clusters of individuals of the same rank obviates the 

 use, as subspecific entities, of such taxonomic groups as variety and form. The use of 

 these terms has been avoided in connection with organisms which, although they have 

 a different appearance, are linked together by series of intermediates which make the 

 boundaries of the taxonomic groups previously used for them entirely unrecognizable, 

 and force us to admit specific unity. Some forms exhibit a set of characters at one end 

 of the cell, and an entirely different set at the other, so that if the ends were considered 

 separately, authors in the past have regarded them as different species. Such variations 

 have been interpreted as phases of the one species, and as a result the number of specific 

 names used has been reduced. The phases are described by the epithets they replace. 

 For example, Rhizosolenia hebetata Bailey, a large and variable species, is found pos- 

 sessing either plain deeply conical valves, or deeply conical valves terminated by a long 

 slender spine. The former is what might be called the type phase, that is, it corresponds 

 with the original description, and the latter is described as the "semispina" phase, as it 

 corresponds with R. semispina Hensen, which has been described as a form of R. 

 hebetata by Gran. Often specimens are found possessing the characters of one species 

 at one end of the cell and characters of the other species at the opposite end. Gran 

 (1904) described such combinations under the name R. hebetata (Bail.) Gran emend. 

 Two forms were described, forma hiemalis Gran, and forma semispina (Hensen) Gran. 

 Under the former Gran placed as a synonym, R. hebetata Bailey, and under the latter 

 R. semispina Hensen. Gran was followed by Hustedt and Lebour. For my own part 

 I do not favour this treatment of the subject, because if the type of R. hebetata Bailey 

 is placed in the synonymy of R. hebetata forma hiemalis Gran, and R. semispina Hensen 

 is placed in the synonymy of R. hebetata forma semispina (Hensen) Gran, we might well 

 ask what is R. hebetata (Bail.) Gran emend. Apparently it has no standing in itself but 

 only in the two forms that compose it. 



Another outstanding example of the application of what might be called the phase 



