DISTINGUISHING IDENTICAL AND FRATERNAL TWINS 



409 



author are probably due to some mistake 

 in the diagnosis of the chorion; and more- 

 over, as he maintains, the percentage of 

 monochorionic births among the total 

 twin-births, according to his estimate, 

 conforms very well to the percentage of 

 monozygotic twins among the whole 

 twin population calculated by the differ- 

 ential method. 



Whether any small percentage of such 

 puzzling cases as mentioned above exists 

 or not among twin births, accurate infor- 

 mation on the foetal membranes is usually 

 lacking for births outside of the hospital. 



Thus, we have still to seek for the 

 criterion for classifying twins in their 

 physical and psychical features. 



PHYSICAL FEATURES 



Of all physical features, our attention is 

 naturally directed first to physiognomy. 

 As Dahlberg remarks (19x6), we have a 

 very sharp sense for discriminating slight 

 details of facial appearance, so that the 

 close resemblance of twins in physiognomy 

 to such a degree that even near relatives 

 can not tell them apart, must involve the 

 identity of their various minute facial 

 features to a very great extent. Thus 

 physiognomy can undoubtedly serve as a 

 good criterion for classifying twins . This 

 method, however, has a drawback in that 

 the judgment of the degree of resemblance 

 depends largely on our subjective sense, 

 and is naturally more or less arbitrary. 

 Many workers, accordingly, seek to rely 

 on more objective anthropological meas- 

 urements and descriptions of other physical 

 characters, such as stature, head form, 

 hair color and form, skin color, blood 

 group, etc. 



Muller (1915), among others, has de- 

 vised a scheme of calculating the a 

 priori chance that given twins are identi- 

 cal on the basis of the data obtained 

 from certain unrelated physical characters 



of the twins and their siblings like stature, 

 hair form, skin color, etc. Thus, accord- 

 ing to him, 



"If there are n sibs altogether (including the twins), 

 a of them falling into class a with respect to a 

 given trait, b into class b, e into class c, etc., the chance 

 that the twins should have been found in the same 

 class is; 



a a — 1 b b — zee — i 



4. 1 h etc. 



n n — 1 w » — 1 n n — z 



If, now, the traits are inherited independently, as 

 they usually will be to a large extent, the chance that 

 two sibs should be in the same class in respect to all 

 of the traits considered is the product of all these 

 chances found in the case of each separate trait" 

 (p. 436). 



And the chance that the given twins are 



p 

 identical is approximately wt^, where 



p stands for the chance that non-identical 



twins should agree in all the traits. The 

 mathematical ground for the formula has 

 been criticized by Miss Burks (192-6) and 

 defended by Muller (192.6). 



It has been remarked by certain authors 

 that identical twins represent the right and 

 left halves of one individual, and that 

 there is often symmetry reversal in some 

 feature or other. The existence of such 

 symmetry reversal is a sign that the twins 

 are monozygotic. Wilder (1904) and New- 

 man (1917) have expressed their view 

 in favor of this idea. Studies have been 

 carried out by some workers to elucidate 

 this interesting question of the asymmetry 

 and symmetry reversal of twins. Handed- 

 ness has been studied by Siemens (19x4b), 

 Lauterbach (19x5), Dahlberg (19x6) and 

 v. Verschuer (19x7), the direction of 

 head whorl by Lauterbach (19x5) and v. 

 Verschuer (19x7), the height of testes 

 in scrotal sac by v. Verschuer (19x7), 

 besides other characters such as the mode 

 of clasping hands and functional superi- 

 ority of one leg examined by Dahlberg 



