94 ON HEREDITY, [II. 



mechanism, which impels him to start forward on hearing a 

 report. We cannot yet determine without more experiments 

 how such an impulse due to perception (' Wahrnehmungstrieb/ 

 Schneider) has arisen ; but, in m}^ opinion, it is almost incon- 

 ceivable that artificial breeding has had nothing to do with it ; 

 and that we are here concerned — not with the inheritance of 

 the effects of training — but with some pre-disposition on the 

 part of the germ, which has been increased b}^ artificial 

 selection. 



The necessity for extreme caution in appealing to the sup- 

 posed hereditary effects of use, is well shown in the case of 

 those numerous instincts, which only come into play once in a 

 life-time, and which do not therefore admit of improvement by 

 practice. The queen-bee takes her nuptial flight only once, 

 and yet how many and complex are the instincts and the reflex 

 mechanisms which come into play on that occasion. Again, in 

 many insects the deposition of eggs occurs but once in a life- 

 time, and 3'et such insects always fulfil the necessary conditions 

 with unfailing accuracy, either simply dropping the eggs into 

 water, or carefully fixing them on the surface of the earth 

 beneath some stone, or laying them on a particular part of a 

 certain species of plant ; and in all these cases the most com- 

 plicated actions are performed. It is indeed astonishing to 

 watch one of the Cynipidae {Rhodites rosae) depositing her eggs 

 in the tissue of a young bud. She first carefully examines 

 the bud on all sides, and feels it with her legs and antennae. 

 Then she slowly inserts her long ovipositor between the 

 closely-rolled leaves of the bud, but if it does not reach exactly 

 the right spot, she will withdraw and re-insert it many times, 

 until at length, when the proper place has been found, she 

 will slowly bore deep into the very centre of the bud, so that 

 the egg will reach the exact spot, where the necessary condi- 

 tions for its development alone exist. 



But each Cynips lays eggs many times, and it may be argued 

 that practice may have led to improvement in this case ; we 

 cannot however, as a matter of fact, expect much improvement 

 in a process which is repeated, perhaps a dozen times, at short 

 intervals of time, and which is of such an excessively complex 

 nature. 



It is the same with the deposition of eggs in most insects. 



