( cxxvii ) 



as a result of our study of these activities is that their com- 

 plexity and perfection have been greatly over-estimated. "We 

 have found them in all stages of development and are con- 

 vinced that they have passed through many degrees, from the 

 simple to the complex, by the action of natural selection. 

 Indeed, we find in them beautiful examples of the survival of 

 the fittest."* 



As long ago as 1889 the present writer had argued that the 

 Lamarckian interpretation of the instincts of Ammophila or 

 Sphex introduced the same difficulty as that alluded to in the 

 discussion of the cocoon-making instinct. It implied a gift of 

 prophecy, a knowledge of what would happen to offspring after 

 the burrow had been sealed and left to its fate.f 



Another powerful argument is derived from the comparison 

 between the instincts which are performed but once and those 

 which are performed many times in a single life. Various 

 elaborate performances are undertaken but once in an insect's 



* See the review of Dr. and Mrs. Peckham's work in "Nature," vol. 

 lix, 1898, pp. 466-468. 



t The argument was used in the ' ' Discussion on Acquired Cliaracters " 

 in Section D of the British Association at Newcastle, Friday, September 

 13, 1889. See Report, p. 620, where, however, only the title of the paper 

 is printed. The following sentences are quoted from the abstract in 

 "Nature," vol. xl, 1889, p. 610 :— 



" With regard to instinct. Dr. Romanes had suggested a difficulty — 

 that was, the instinct of certain wasps to sting and paralyze the nerve 

 centres of their prey. But it must be remembered that the benefits 

 arising from this instinct were felt not by the wasps themselves, but 

 by their progeny." 



In " Proc. Boston Soc. Nat. Hist.," vol. xxvi, 1894, p. 392, the argu- 

 ment is stated in greater detail as follows : — 



" The wasp-like insect has no opportunity of learning by experience 

 because it can never know whether the larva stored up is a failure or a 

 success. If the larva had not been stung, or, accepting the received accounts, 

 had been stung in the wrong place, it would struggle and perhaps kill the 

 young grub ; or dying of starvation it might dry up and be useless as food. 

 But the Hymenopteron never goes back to inquire. It makes all the 

 difference to the young grubs whether the food provided for them is in an 

 appropriate condition or not, but it makes no difference whatever to the 

 parent insect. The latter seals up the chamber in which its eggs have 

 been laid and never opens it again ; it has no chance of noting the failure 

 or the success of the food it has provided. It is clearly a case like that of 

 the cocoon, which cannot be explained on the Lamarckian theoiy and 

 must be explained on the Darwinian. And this latter interpretation is 

 easy ; those insects which possessed the nervous mechanism impelling 

 them to provide food in an appropriate condition gave to their offspring 

 the opportunity of surviving and inheriting the same instinct ; while 

 others, impelled to perform less efficient actions, were thereby cut off from 

 any representation in the next generation. " The passage has been slightly 

 modified. 



