C. L. PROSSER 341 



the tolerated limits. Most survival tests are made with adults 

 rather than with young, a procedure which is easier but perhaps 

 less meaningful for speciation, since embryos are frequently more 

 sensitive to stresses. Despite the fact that survival data must be 

 qualified according to acclimation history, time of observation, 

 and rate of application of the stress, the extremes to which ani- 

 mals are subjected in nature are more important than are average 

 environmental values. 



Reproduction. Ability to reproduce measures survival over 



Fig. 1. Diagrams representing relation between an internal state I and 

 an environmental stress O. a x and a 2 represent environmetal limits of two 

 acclimation levels corresponding to survival limits s x and s 2 of internal 

 variation. Broken lines represent regions where only brief survival is pos- 

 sible. (A) Diagram for a conforming (adjusting) animal. (B) Diagram for 

 a regulating animal. From Prosser (1955). 



a complete life cycle and approaches the natural situation better 

 than survival of adults only. Unfortunately, rearing animals in 

 the laboratory is possible for relatively few species, and the best 

 comparisons are of reproduction in populations occurring in 

 different ecological situations and acclimated by transplanting to 

 different field conditions. More significant than ability to repro- 

 duce is rate of growth of populations, and this can be measured 

 readily in the laboratory only for a few species of small animals, 

 particularly protozoans and insects. 



Measurement of an Internal State as a Function of External 

 Stress. There are two basic patterns of physiological adaptation 



