BY R. J. TILLYARD. 25 



Species, Aj, Ag, A^, the contour will not be aifected unless trvo 

 or more of the forms occur in a single locality. But as nearly all 

 the argument between " splitters " and " lumpers" occurs about 

 " geographical races," it follows that in such cases no alteration 

 of the contour is effected by a change of opinion, since the species 

 in question can only score " one " in each locality in which it 

 occurs. 



Local varieties, known to be produced as the offspring of a 

 definite species, should on no account be included as "species." 



V. Application of the Method. 



On the map of the area to be studied, each locality from which 

 records are obtainable should be marked down. Against each, 

 the number of species (of the group in question) occurring in 

 that locality should be written. Contour lines are then to be 

 drawn as free curves enclosing in tnrn all those localities possess- 

 ing the same number of sjjecies. In the simplest cases (where no 

 lacunae or breaks occur), the result will be as follows : — 



Between the outermost contour-line (1) and the next (2) will 

 lie all those localities in which only one species occurs. (N.B. — 

 It is important to notice that this is not necessarily the same 

 species for all these localities). 



Between contour-lines 2 and 3 will lie all those localities pos- 

 sessing ttvo species (again, not necessarily the same two species). 

 And so on. Finally, the nth or highest contour-line will be 

 either a closed oval, or possibly a series of closed ovals, of com- 

 paratively small extent, enclosing those few localities in which 

 the highest total of n species occurs (again, be it noted, not 

 necessarily the same n species in every locality within an oval). 



Where the records are not sufficient, continuity or discon- 

 tinuity may be assumed provisionally according to the evidence 

 available. To give an example : — A species may be recorded 

 from Sydney, Newcastle, Richmond River, Tweed Heads, Bris- 

 bane, Rockhampton. In such a case, it may reasonably be 

 assumed that it occurs along the whole coast-line from Sydney to 

 Rockhampton, because the conditions known to exist between 

 these points naturally suggest its occurrence throughout. But, 



