24 THE BIOLOGY OF TWINS 



embryonic connections. It is hardly Kkely that two 

 outgrowths could turn around or reverse their positions, 

 at least not until the umbilical cords became elongated. 

 Even armadillo quadruplets which are in separate amnia 

 are found in advanced stages lying in positions such 

 that, should fusions occur between adjacent individuals, 

 they would unite to form diplopagi strictly bilaterally 

 symmetrical with reference to one another. 



It is highly probable that certain cases of doubling 

 in head and posterior regions of human twins are due 

 to disturbances in the process of concrescence of the 

 right- and left-hand component of such embryonic 

 primordia as the neural groove and the ventral body 

 suture. In various vertebrates all sorts of partial dou- 

 bling may be produced by experiment. For example, 

 I have worked with certain strains of hybrid fish in 

 which double-headed and double-tailed fish have been 

 the result of abnormal developmental conditions. It 

 may well be true, then, and probably is, that not all 

 cases of doubhng in human embryos are due to the 

 same type of cause; some may be due to fission and 

 others to fusion. Where we have so little evidence on 

 the embryological side, it seems unwise to speculate 

 further upon the causes of twinning and doubling in man. 

 The only real clue as to the mode of twinning in 

 man comes from a study of polyembryonic development 

 in the armadillo. For various reasons it is beheved 

 that the process of monozygotic twinning in man i? 

 essentially the same as that of the armadillo; hence 

 a detailed study of the facts revealed by a study of 

 the , armadillo is the nearest approach possible today 

 to a direct study of human twinning. 



