VARIATION AND HEREDITY IN TWINS 171 



similarity between the main patterns of the two individ- 

 uals and a difference in the exact detailed exi)ression of 

 the pattern in terms of integumentary units, so in 

 armadillo quadruplets there is generally a pattern 

 (double band or double scute) in a similar region of 

 the armor in all four individuals of a set, but there 

 may be a considerable difference in the number and 

 distribution of the integumentary units employed in 

 the expression of the pattern. The pattern (doubling) 

 may be expressed in a large number of units (scutes) 

 in some members of a quadruplet set and in a small 

 number (sometimes only one) of units in others. It 

 may be expressed unilaterally in some and bilaterally 

 in others. Or, finally, the pattern may be expressed in 

 some members and totally suppressed in others. 



When, in his examination of same-sexed twins, 

 Wilder encountered cases in which certain patterns 

 were not sufficiently identical to meet his preconceived 

 ideas of what duplicate twins should be, he concluded 

 that they were fraternal twins (dizygotic). In the 

 light of what I have found in armadillo quadruplets, 

 which are unquestionably monozygotic, it does not 

 seem safe to exclude from the category of monozygotic 

 twins those that fail to show identity of pattern; the 

 same practice, if applied to armadillo quadruplets, 

 would lead to grave error. It is probably safer to say 

 that som€ monozygotic human twins, like some sets 

 of armadillo quadruplets, are nearly identical, while 

 others, hke various quadruplet sets, may dilTer materially 

 from each other. 



To me it appears almost certain that Wilder's twins, 

 Nos. VII, X, and XIII, are monozygotic (duplicates), 



