146 HEREDITY AND ENVIRONMENT 



gets the failures of a rule and remembers the 

 times when it holds good it is possible to be- 

 lieve in the influence of food or temperature 

 or age, of war or peace or education on the 

 relative numbers of the sexes, or on almost any 

 other thing. By statistics it has been shown 

 that each of these things influences the sex 

 ratio, and by more extensive statistics it has 

 been proved that they do not. 



This was the condition regarding the causes 

 of sex determination which prevailed up to 

 the year 1902. Immediately preceding that 

 year it had been found that two kinds of sper- 

 matozoa were formed in equal numbers in cer- 

 tain insects; one of these kinds contained a 

 peculiar "accessory" chromosome, and the 

 other lacked it. The manner in which these 

 two types of spermatozoa were formed had 

 been carefully worked out by several investi- 

 gators without any suspicion of the real sig- 

 nificance of the facts. It was shown that an 

 uneven number of chromosomes might be 

 present in the spermatogonia of certain in- 

 sects and that when maternal and paternal 

 chromosomes united in pairs in synapsis one 



