279 



3. The variation in the anal rays and dorsal spines are determinate with the 

 latitude, the southern specimens having a slightly larger number of rays and 

 spines. 



4. The color-pattern variations are determinate, varying through definite 

 stages from a simple to more complex pattern. 



In Table A and B are given the data on the anal rays and dorsal spines. 

 The localities are arranged in the order of their latitude from north to south. 

 From these we see that there is both an increase in the average number of rays 

 and spines and in the number that prevails in each case from north to south. In< 

 the anal fin 10 is the prevailing number north, and 11 and 12 south, of the Ohio 

 Biver. Fourteen and fifteen are the prevailing number of dorsal spines in the 

 north and 15, 16 and 17 in the south. 



TABLE A. 



LOCALITY. 





a « 



« a 



Q> O 



ft"" 1 



§ ap3 



0,1-H 



™x. 



CD CO _ 



-^ a >-> 



CD C3 



a« 



QJ CI 



°'* . 



CD CO CO 



Torch Lake, Mich 



Cedar Eapids, Iowa 



White Biver, at Indianapolis 



Gosport, Ind 



Bean Blossom, Ind 



Bushville, Ind 



Wild Cat Creek, Ind 



Pike Creek, Ind 



Illinois 



Nipisink Lake, 111 



Monongahela Biver 



Hartford, Ky 



Green Biver, Greensburg, Ky 



Little Barren Biver, Osceola, Ky 



Little South Fork Cumberland Biver, Wayne 



County, Ky 



Eagle Creek, Olympus, Tenn 



Obeys Biver, Elizabethtown, Tenn 



Watauga Biver, Elizabethtown, Tenn 



North Fork Holston Biver, Saltville, Va 



Eureka Springs, Ark 



Chocola Creek, Oxford, Ala 



San Marcos Springs, Tex 



7 

 1 

 1 

 5 

 17 

 1 

 1 

 2 

 1 

 2 

 1 

 4 

 3 

 4 



1 

 2 

 13 

 2 

 1 

 1 

 4 

 2 



1(H- 



12 



10 



10 



10fV 



10 



11 

 11 



10 



10?, 



10 



m 



lOf 



11 



11 

 11 

 HA 



12 



11} 

 11 



