C. L. PROSSER 359 



species of the butterfly Colias, eurytheme on alfalfa and philodice 

 on red clover ( Hovanitz, 1949 ) . 



Other environmental factors which may be important as iso- 

 lating agents are such sensory stimuli as light and mechanical 

 stimuli. For example, blind races of the Mexican characin, Ano- 

 ptichthys jordani, inhabit dark caves where lack of vision is not a 

 disadvantage ( Breder, 1943 ) . Wind and water currents, also sub- 

 strate composition, limit the distribution of a variety of animals. 

 Biotic Selection. Related to food plant preference as an iso- 

 lating mechanism is the selection of host organism or of symbiont. 

 Termites which are very similar morphologically are clearly dis- 

 tinguished by the species of staphylinid beetle living in their nests 

 (Emerson, 1935). Ascaris lumbricoides is morphologically similar 

 in man, pig, and monkey, yet attempts at reciprocal infection 

 have been unsuccessful ( Faust, 1949 ) ; these ascarids may be true 

 sibling species. Probably the sibling species most difficult to un- 

 derstand are those of the Anopheles maculipennis complex. Eight 

 different strains are recognizable by their selection of man or 

 other mammals as blood sources, also by their ability to develop 

 in saline water. These strains are morphologically similar except 

 for minor differences in egg structure. However, laboratory 

 breeding experiments demonstrate that in reality the "strains" are 

 sibling species (Hackett and Missiroli, 1935; Bates, 1949). An 

 American mosquito, Culex pipiens, has two strains, pipiens, which 

 for egg production requires a blood meal, and molestus, which 

 can produce eggs without it (Kitzmiller, 1953). 



Serological specificity distinguishes closely related groups and 

 reveals incompatibility between more distant groups. "Blood 

 types" have been analyzed in detail in man, in pigeons, and in 

 cattle. In man, certain serological types, also certain hemoglobin 

 types, are correlated with other genetic factors, and each type is 

 carried at a level characteristic of the ethnic group ( Boyd, 1950 ) . 

 Present suggestions of linkage of blood type with incidence of 

 certain chronic diseases indicate the need for greatly expanded 

 medical statistics. 



Behavior. Fully as important as physical and chemical factors 

 in maintaining (possibly not in initiating) isolation between pop- 



