280 THE COLOUES OF ANIMALS 



this hinder part from side to side, and with it the 

 antenna-like appendages. This movement also adds 

 to the general effect. 



When the larva is slightly irritated, the position is 

 often imperfectly assumed at first, but as the irritation 

 is repeated and increased, the animal adds the various 

 details which go to make up the terrifying attitude in 

 its most perfect and elaborate form. 



Experimental proof of the protective value of the 

 terrifying- attitude in the Lobster caterpillar 



I offered two of these larvae to the marmoset, and 

 the results proved the importance of conducting such 

 experiments with the greatest care, if reliable con- 

 clusions are to be obtained. The marmoset knew 

 that my boxes contained insects, and was always very 

 keen and excited at the sight of them. When the box 

 containing one of the ' Lobsters ' was opened, the cater- 

 pillar was seized and devoured before it had time to 

 alter its position, and before the marmoset could have 

 had the chance of being intimidated. The second 

 caterpillar was placed on the table and made to assume 

 its terrifying attitude, and then the marmoset was 

 allowed to approach it. Although a caterpillar of the 

 same size had just been eaten without the slightest 

 hesitation, the marmoset was much impressed by the 

 alarming sight, and only ventured to attack after the 

 most careful examination, and even then in the most 



