EFFECTS OF A RISING TEMPEEATDEE. 117 



regions or of the eternal snow-fields of high mountain peaks. 

 In one single case only can we assert with certain y that the 

 winter colouring of an animal may be referred to the direct in- 

 fluence of the reduced temperature in autumn. Professor Weis- 

 mann in Freiburg, to whom we are indebted already for many 

 facts established by his elegant experiments, has proved that 

 two varieties of butterfly, long regarded by entomologists as 

 distinct species, are in fact only the summer and winter forms 

 of the same species of Vanessa {Vanessa prorsa-levana), for 

 he succeeded in rearing the winter variety ( Vanessa levana) in 

 the summer season, and from a summer brood, by keep- 

 ing the air in which the caterpillars and pupse lived at an 

 artificially lowered and regular temperature.^" It is much to 

 be wished that zoologists wou^d more frequently carry out 



Fig. 29. — BesoHa glacialis, the G-lacier Flea. 



similar experiments; for if this were done, I have no doubt that 

 a far more extensive influence of cold on. animals would soon be 

 recognised, and its limits more accurately defined than is at 

 present possible. 



II. The influence of a rising temperature on animal life. — 

 Variations in temperature below the freezing-point of the fluids 

 contained in the body can obviously have no effect on those 

 animals, or on their peculiarities, which either die when the 

 cold reaches that point, or whose vitality becomes latent. It 

 will matter little to a perfectly frozen frog whether it was 

 frozen at 5° below zero or at 10° or 12° above. It is not until 

 the water surrounding the creature, and the fluids contained in 

 its tissues, recover their fluidity under a thaw, that any further 

 rise of temperature can affect its vital activities. For although 

 many animals can live on snow and ice, or even in ice — as the 



