ORGANIC EVOLUTION 



273 



ing off the classes-upon the ordinates and the frequencies 



upon the abscissae. 



Then, if this work be done by a class, let the totals 



of all the individual 

 counts be represented 

 i n another curve, 

 plotted in another 

 color upon the same 

 square; this will, on 

 account of the greater 

 numbers, more truly 

 represent the normal 

 variation for the 

 species, and it should 

 be a closer approxi- 

 mation to the sym- 

 metrical and balanced 

 curve of distribution 

 of error. 



The record of this 

 study will consist in: 



i) A drawing of a 

 variant showing the 

 normal condition for 

 the species, labelled 

 with the name, and 

 showing clearly the 

 parts counted and 

 plotted. 



2) The individual and collective curves of variation. 

 Mutation. Variations are not all so light and inconstant. 

 Figure 167 shows a variant of the common linaria (L. vul- 

 garis, "butter and eggs"), the ordinary flowers of which are 



FIG. 167. A probable mutant of Linaria (L. 

 vulgaris), "butter-and-eggs." 



