December 9, 1910] 



SCIENCE 



845 



spots." As a matter of fact, in the great ma- 

 jority (about five sixths) of cases of fusion 

 the spots are well joined. 



In the following tabulation of condition of 

 the varying color pattern in the species (on 

 the Stanford University campus) in differ- 

 ent years, series of approximately 1,000 are 

 used. That a series of 1,000 individuals col- 

 lected from one locality at one time fairly 

 represents, in the variation revealed by its 



wise noted were collected on flowers in ap- 

 proximately the same place on the campus. 

 All the lots were collected in October. 



The following tabulation gives the data for 

 all the series examined since 1895. 



That the relative frequencies of the prin- 

 cipal two pattern conditions may be more 

 readily obvious I present in the following 

 table the percentages of these types for each 

 of the lots studied. 



members, the actual variation conditions of 

 the species in this locality, as regard both 

 kinds of variation and frequency of these 

 kinds, is proved by repeated tests made by 

 examining and tabulating successive thou- 

 sands taken at random from the same place 

 at the same tiiue and finding a practical ident- 

 ity in these separate lots. Indeed, series of 

 ■500 gave practically approximately the same 

 proportions as those of 1,000. But the larger 

 number is the safer. All the lots not other- 



' Lateral fusion of anterior or posterior spots; 

 longitudinal fusions, etc.; always includes some 

 eases of lateral fusion of middle spots accom- 

 l-anying fusions. 



