191 i] Muttkowski, Studies m Tetragoneuria. 119 



the moment, it was immediately apparent that for the rest of the 

 Tctragoncurice the color development showed a progression from 

 south to north : Specimens from North Carolina, Ohio, Indiana 

 and New Jersey showed much less color development than speci- 

 mens from New York, Maine, Ontario and Wisconsin. This 

 phenomenon was obviously at variance with the apparent develop- 

 ment of the semiaquea series. 



Here, then, existed a state, North Carolina, one of the most 

 southern places from which Tetragoneurice had been collected, 

 in which the color had reached a higher development in the wing 

 of the species than along the line of its regular geographical 

 distribution. Here, in the compass of a few miles, the same 

 results were achieved which otherwise necessitated a broad geo- 

 graphical expanse of approximately a thousand miles. 



To co-ordinate this apparent fact with the real fact of color 

 development from south to north seemed an impossibility. 



It was about this time (March) that Mr. Brimley sent me an 

 extensive series of species from North Carolina. These together 

 with other material from North Carolina already at hand, formed 

 the basis of a prolonged study in which the final conclusion was 

 reached that semiaquea is a distinct spedies. 



Distributing this material among the color divisions c to i, 

 as noted under cynosura, it was found that they easily divided 

 into two lots: (a) color not extending beyond triangle and (b) 

 color reaching the nodus. 



(a) Among the first lot the large majority fell under the 

 color type C2 — the color reaching the first antecubital, two speci- 

 mens came under the intermediates d-e, while but four fell under 

 e, and none at all under e-f, f and g. Considering this lot in the 

 light of the evidence shown by material from other regions the 

 large majority of forms of C2 was conclusive: it marks the 

 typical southern form of cynosura, the first step in the color 

 progression toward the north as found in this species. 



(b) The second lot showed the three forms il, 12 and i3 as 

 before indicated. 



A further result was remarkable: only two other specimens, 

 from New Jersey and Massachusetts, reached the high color 

 development shown by the North Carolina forms. Furthermore, 

 the nearest approach to semiaquea — the forms f and f-g of simu- 

 lant — occur onlv in northern states (Wisconsin, Maine, Ontario 



