ni MARKINGS OF CATERPILLARS 



73 



Weismann discovered in his caterpillars a regular series of 

 steps in the marking from longitudinal lines °to transverse 

 lines and spots, and came to the conclusion : '' That among 

 species which are ornamented with oblique streaks or with 

 spots there are many whose young stages are streaked length- 

 wise, but the converse never occurs; never do the young 

 caterpillars exhibit spots or oblique streaks when tlie adult 

 caterpillar has only longitudinal streaks. The first and most 

 ancient marking of the Sphingida3 caterpillars was therefore 

 the longitudinal striping." A derivation of the oblique from 

 the longitudinal striping is not demonstrated; both occur 

 together, but the former appears later than the latter, and re- 

 mains when this has vanished. 



I thought it possible that the longitudinal streaks broke 

 up first into small spots and dots, and that these, spots 

 and dots formed the oblique streaks. If so, complete agree- 

 ment with my laws of marking would be proved in this 

 case. 



In caterpillars also, as already mentioned, the new charac- 

 ters arise mostly at the posterior end of the body. 



Weismann endeavours, as before remarked, to explain the 

 markings of caterpillars as adaptations to the conditions of 

 the environment. Nevertheless, he comes to the conclusion 

 that these markings have phyletically developed extremely 

 gradually, according to certain laws, and in quite definite 

 directions. And he further says : " The evolution of the 

 species of Deilephila shows that the evolution of the marking 

 follows throughout a certain law, that it proceeds in all 

 species in the same manner. All the species seem to steer 

 towards the same point, and this gives the impression that there 

 is an internal law of evolution which, like an impelling force, 

 determines the future phyletic modification of the species." 

 This conclusion seems to be supported also by the fact that in 

 caterpillars there is a tendency for the same characters to be 



